» lilil iill W ! ! ll CONTRIBUTIONS FriOM OEPARTMENT OF GEOlOGY ^PRINCETON UNIVERSITY CONTRIBUTIONS TO PAL^ONTOLOGY I UPPER CRETACEOUS FLORAS OF THE ROCKY MOUNTAIN REGION I: STRATIGRAPHY AND PAL^ONTOLOGY OF THE FOX HILLS AND LOWER MEDICINE BOW FOR- MATIONS OF SOUTHERN WYOMING AND NORTHWESTERN COLORADO By Erling Dorf Princeton University With nineteen plates and 8 text figures 7- [Preprinted from Caenegie Institution of Washington Publication No. 508, pages 1 to 78, December 12, 1938] CONTRIBUTIONS TO PAL/EONTOLOGY I UPPER CRETACEOUS FLORAS OF THE ROCKY MOUNTAIN REGION I: STRATIGRAPHY AND PAL/EONTOLOGY OF THE FOX HILLS AND LOWER MEDICINE BOW FOR- MATIONS OF SOUTHERN WYOMING AND NORTHWESTERN COLORADO By Erling Dorf Princcton University With nineteen plates and 8 text figures [Issued December 12, 1938] CONTENTS PAGE Introduction 1 Present Physical Conditions and Vegetation 2 Stratigraphy 4 Southcrn Wyoming 4 Northwestern Colorado 7 Local Sections and Maps 9 Composition of the Fiora 22 Systematic Relationships 22 Quantitative Analysis 23 Leaf Characters 25 Past Climatic Conditions as Indicated by the Flora 27 Correlation 31 Summary 36 Systematic Palteobotany 38 General Procedure 38 Synonyms and Changes of Names 40 Descriptions 43 STRATIGRAPHY AND PALAEONTOLOGY OF THE FOX HILLS AND LOWER MEDICINE BOW FORMATION OF SOUTHERN WYOMING AND NORTHWESTERN COLORADO INTRODUCTION Iii view of the controvcrsial naturc of thc Lancc and rclati>d floras in connection with ]\Iesozoic-Cenozoic boundary problems, it is rathcr disconccrting to find that neither the Lance nor the closely related Medicinc Bow floras have ever been ade- quately studied or described. It is the purpose of this paper to outhne the stratig- raph}' and palseontology of the Medicine Bow formation, and to describe and discuss the fossil plants collected from this formation ckiring the past two years. The study of the Lance floras of Wyoming, Montana, and the Dakotas is in progress and will form the basis of subsequent papers. While my interest is primarily in the succession of Upper Cretaceous floras of the West and their relation to succeed- ing Cenozoic floras, it is hoped that the rcsults of these studies will help to harmonize the conflicting lines of palseontologic and stratigraphic evidence regarding the Mesozoic-Cenozoic boundary in the Rocky Mountain region. For many years, confusion has arisen from the practice of including all the coal- bearing formations of the Upper Cretaceous and lower Tertiary in the all-embracing "Laramie Group"; as a result, few of the contained floras could be adopted as standards for comparison in other fields. As in all other branches of stratigraphic palseontology, it is necessary first to establish a standard fauna or flora whose age is definitely known from other lines of evidence. Thereafter, the age of other faunas or floras can be determincd by comparison and contrast. The Medicine Bow flora is admirably suited to serve as such a standard because its age can be rather defi- nitely determined by its relation to both invertebrate and vertebrate faunas of known age. The study of the Medicine Bow flora is part of a program on Upper Cretaceous floras of the Rocky Mountain region which I have vmdertaken under the joint auspices of the Carnegie Institution of Washington and Princeton University. The distribution of the 6 localities f rom which the fossil plants were obtained is indicated in figure 1. At each of these locahties the sections were carefully measured by compass-traverses, and the exact stratigraphic position of each plant-bearing horizon was determined with reference to a well-established and diagnostic faunal zone of the underlying Fox Hills formation. Sections were measured by A. P. Conway, Dean Hill, and M. L. Rittenhouse, students at Princeton University, who also identified most of the invertebrates coUected. The invertebrate identifications were checked by Dr. John B. Reeside of the U. S. Geological Survey. Friendly encouragement and constructive criticism have been freely given by Dr. Ralph W. Chaney of the University of Cahfornia. Yaluable suggestions in various phases of the study have been made by Dr. Roland W. Brown of the U. S. Geological Survey, and Dr. S. H. Knight of the University of Wyoming. Herbarium material has been 1 2 CONTRIBUTIONS TO PAL^ONTOLOGY generously loaned by the New York Botanical Garden. Competent field assistance was rendered by Beverly Wilder of Antioch College, H. H. Sharkey of Princeton University, and Ruth K. Dorf. FlG 1 — Index map, showing location of Localities P. 371 to P. 376. PRESENT PHYSICAL CONDITIONS AND VEGETATION Of the 6 locaUties studied, 4 are situated in southern Wyoming and the remain- ing 2 in northwestern Colorado, as shown in figiu-e 1. Both areas are essentially alike in physical and climatic conditions and differ only in details with respect to the type and distribution of vegetation. These areas are included in the high, barren intermontane basins of the Rocky Mountain system. Their average elevation ranges from approximately 5000 to 7000 feet. The topogi-aphy is sub-maturely dissected, and largely controUed by the differential erosion of variously dipping strata of unequal hardness. The southern Wyoming region Ues in the drainage area of the North Platte and Medicine Bow rivers, which are the only permanent streams. The Yampa River drains the area in northwestern Colorado. Thc narrow flood plains of these streams are features of primary importance in the distribution of modern vegetation. On the basis of moisture and temperature conditions these areas are near the eastern edge of the cool, arid cUmatic province.^ Normal annual precipitation > Livingston and Shreve, Carnegie Inst. Wash. Pub. No. 284, 380, 1921. FOX HILLS AND LOWER MEDICIXE BOW 3 avcragcs bctwcen 10 and 15 inchcs, little of which falls during thc growing season, which usually lasts from 80 to 110 days in the summer months, June to Septcmher. The mcan annual tcmpcrature is rather low, avcraging from 40 to 45° F. During the short growing season, however, tempcratures of over 100° F. are not unusual. Koppen and Geiger ' place the areas in the Dry, Cold, Steppe climatc, BSKW. Ecologically, the vcgetation in thc areas of both southern Wyoming and north- westcrn Colorado may be divided into two distinct units: the sagebrush-grassland association of the dry ridges and divides, and the poplar-willow association of the moist, narrow flood plains of the permanent streams. In the following list of com- mon species, those present in southern W'yoming, but not in northwestern Colorado, are marked with an asterisk (*) ; those in northwestern Colorado, but not southern Wyoming, with a dagger (f) ; the unmarked species are prevalent in both areas.- Ridges and Divides (Sagebrush-Grassland Association) : Trees and shrubs — Artemisia tridentata Nuttall — Black sagebrush Atriplex nuttallii S. Watson — Salt sage Chrysothamnus graveolens (Nuttall) Greene — Rabbit brush Eurotia lanata (Pursh) Moquin — White sage or Winter fat Grayia spinosa (Hooker) Moquin — Hop sage Juniperus spp. — Cedars Kunzia tridontata (Pursh) Sprengel t Pinus eduHs Engelmann — Pinyon pine Sarcobatus vermiculatus (Hooker) Torrey — Greasewood Herbs — Agropyron spicatum (Pursh) Scribner and Smith — Wheat grass Bouteloua ohgostachya Torrey — Short grass Bulbilis dactyloides Rafinesque-Schmaltz — Buffalo grass Elymus sp.? — Rye grass Opuntia spp. — Prickly pears Stream Flood Plains (Poplar-Willow Association) ; Trees and shrubs — Amelanchier alnifolia Nuttall — Service berry Cercocarpus ledifoHus Nuttall — Mountain mahogany t Crataegus rivularis Nuttall — Black haw Populus angustifolia James- — Narrow-Ieaved cottonwood * Populus sargentii Dode — Broad-leaved cottonwood * Ribes longiflorum Nuttall — Flowering currant Ribes spp. — Gooseberry Rosa sp.? — Wild rose t Salix amygdaloides Anderson — Willow Salix longifolia Miihlenberg — Willow Shepherdia argentea Nuttall — Buffalo berry t Symphoricarpus oreophilus Gray — Waxberry ' Koppen and Geiger, Klimakarte der Erde (Gotha, Justua Perthes), 1928. ' Detailed observations on distribution are given in Cary, M., N. A. Fauna, vol. 33, 1911, and vol. 42, 1917. STRATIGRAPHY SOUTHERN WYOMING The Medicine Bow formation is an alternating succession of shales, sandstones, and coal beds resting conformably on the Lewis shale in the Hanna and Carbon basins of south central Wyoming. The first comprehensive report on this region was by Veatch,^ wlio designated the formation as "Lower Laramie" on the basis of stratigraphic position and "the absence of marine fossils." For essentially the same beds Bowen ^ subsequently proposed the name Medicine Bow formation, which has since become generally accepted. Regarding the fossils found in the formation, Bowen wrote as foUows: "The formation contains remains of fresh and brackish water invertebrates, land plants, and bones of vertebrates. The plants are regarded by F. H. Knowlton as of the same age as the plants of the Laramie of the Denver Basin. The invertebrates are considered by T. W. Stanton to belong to the fauna of the Lance formation. The bones belong in part to the ceratopsians, but no specimens have been found that are sufficiently diagnostic for even generic determination." In a later report on the same general region, Dobbin, Bowen, and Hoots ^ described the formation in more detail and included within its lower limits the " sandstonc beds with a marine fauna of Fox Hills type." It is clearly evident from both Veatch's and Bowen's statements as quoted above that this marine sandstone had previously been excluded from the Medicine Bow formation and regarded as the upper portion of the Lewis shale. The various interpretations of the late Cretaceous and early Tertiary deposits of the region are shown in figure 2, which is compiled from the three reports cited. The writer's interpretation of the section is presented in the right-hand colimm of figure 2. The recognition of the Fox Hills formation between the Lewis and the Medicine Bow is based upon the occurrence of the Sphenodiscus zone, which is of restricted Fox Hills age,^ at numerous locaUties in the region, and upon the marked changes in the aggregate lithology of the sequence. Veatch apparently collected no fossils from these beds, nor did he mention any changes in Uthology in the upper portion of the LeM'is formation or the lower portion of the "Lower Laramie." Bowen^ reported the presence of a persistent gray sandstone near the top of the Lewis shale, but did not mention the massive, coarse, concretionary sandstones near the base of the Medicine Bow formation. He apparently collected no fossils from these units at that time. The subsequent report by Dobbin, Bowen, and Hoots recordcd characteristic Fox Hills species including Sphenodiscus lenticularis (Owen) Meek near the top of their Lewis formation, and "a few marine beds with in- vertebrate fossils of Fox Hills types intercalated in the basal part" of the Medicine Bow formation.'^ Although they did not recognize the Fox Hills formation as a ' Veatch, A. C, Coal Fields of Easl-cenlral Carbon County, Wyonnng, U. S. Gcol. Survey Bull. 316, 244-200, 1907. ^Bowen, C. F., Stratigraphy of thc Hanna Basin, Wyoming. U. S. Geol. Survey Prof. Paper lOS-L. 227-235, 1918. ' Dobbin, C. E., Bowen, C. F., and Hoots, H. W., Geology and Coal and Oil Kesources of thc Hanna and Carbon Basins, Carbon County, Wyoming, U. S. Gcol. Survey Bull. 804. 9. 23, 1929. * Reeside, J. B., Jr.. written communication, May 9, 1935. ' Bowen, C. F., op. cit., 228, 229. « Dobbin, C. E., Bowen, C. F., and Hoota, H. W., op. cit., 21, 22. 4 FOX HILLS AND LOWER MEDICINE BOW unit in cither the text or the maps, these authors suggested its presence, as shown in the foUowing quotation :' "The Fox Hills sandstone of eastern Wyoming is almost certainly represented in the uppcr- most part of the Lewis shale, [and] in the lower 500 feet of the Medicine Bow formation. . . ." Veatch, 1907 Bowen, 1918 System Group Formation Thicknesa (feet) Sj-stem Formatioa Thicknesa (feet) Tertiary North Park 4,500 North Park 0^00 ± 7,000± 6,500± Fort Union 1,200± Tertiary Hanna — Unconformity — Ferris Laramie hitherto so-called " Upper Laramie " 6,800 ± Tertiary? Cretaceous Cretaceous " Lower Laramie " 6,500 Medicine Bow 6,200 ± Montana Lewis 1,800- 3,000 Lewis 3,300 ± Mesaverde 1,500- 3,200 Mesaverde 2,700 Dobbin, Bowen, and Hoots, 1929 Present Interpretation Age Formation Thicknesa (feet) System Formation Thickness (feet) Miocene? North Park 0-400* 7,000* Miocene? North Park 0-400* Eocene Harma Eocene Harma 7,000* Paleocene Ferris 6,500* Ferris 6,500* Upper Cretaceous Eocene? Medicine Bow 4,000- 6,200* Medicine Bow 4,000- 6,000* Upper Cretaceous Fox HiUs 375-400 Lewis 3,300* Lewis 3,000* Mesaverde 2,200- 2,700* Mesaverde 2,200- 2,700* * Approximate. FiQ. 2 — Late Cretaceous and early Tertiary formations in the Hanna Basin, Wyoming. Since this report Dobbin and Reeside have described several sections in this region in wliich they encountered distinctive faunas of the Fox Hills type in the transition beds between the marine Lewis shale and the non-marine IMedicine Bow formation." The Fox Hills invertebrates, including the diagnostic Sphenodiscus ' Dobbin. C. E., Boweu, C. F.. and Hoots, H. W., ibid., 23. ' Dobbin, C. E., and Reeside, John B., Jr., The Conlact ofthe Fox Hilla and Lance Formations, U. S. Geol. Survey Prof. Paper 158-B. 21, 22, 1929. 6 C O N T R I B U T I O N S TO PAL^ONTOLOGY lenticularis, were shown to come from strata included within the basal part of the Medicine Bow formation and not from the sandstones within the upper part of the Lewis shale. The wTiter's field studies and coUections, which are later described in detail, are in agreement with these observations. It is clearly evident, therefore, that sediments of Fox Hills age form a part of what has hitherto been regarded as the basal portion of the Medicine Bow formation. The recognition of the Fox Hills formation in the region must then depend on whether or not it can be distinguished as a separable lithologic unit. The foUowing criteria for the recognition and de- Hmitation of this formation in northeastern Colorado have recently been pubUshed by a committee of the Rocky Mountain Association of Petroleum Geologists ^ after a field conference with Dr. J. B. Reeside, Jr.: "The base of the Fox Hills formation shall be considered as the horizon below which the section is predominantly gray marine clay shales and sandy shales of Pierre age, and above which the section changes rapidly to a bu£f to brown sandstone containing numerous large gray to brown, hard, sandy concretions. This lower concretionary member is commonly overlain by a series of light gray to brown sandstones and sandy shales. "The top of the Fox Hiils formation shall be considered as the horizon above which the section is composed predominantly of fresh and brackish-water deposits accompanied by coals and lignitic shales, and below which it is predominantly marine." The appUcation of these criteria to the basal 400 ± feet of the Medicine Bow formation has recently been made by Conway in an unpubUshed report.^ He has pointed out that aU of the above requirements are exactlj^ fulfiUed in the sections which he studied near Corson Ranch, near Walcott, and near RawUns in this area. It is interesting to note that the lowest coal beds in the sections are invariably at the horizon which separates the underlying marine faunas from the overlying fresh and brackish-water faunas, and which also separates the massive, more persistent sandstones below from the tliinner, lenticular sandstones above. For practical purposes of mapping, therefore, the lowest coal beds can be regarded as the base of the Medicine Bow formation. The horizons which have yielded plant remains aU Ue within the basal 1300 feet of the Medicine Bow formation, as here redefined. Each horizon has been ac- curately located with reference to the characteristic Fox HiUs fauna, of which the ammonite Sphenodiscus lenticularis is regarded as the most trustworthy and diagnostic index species. It is certain, therefore, that the evidence from strati- graphic position and from marine invertebrates indicates post-Fox HUls age for the plant-bearing horizons. Their pre-Fort Union age is equaUy clearly indicated by the presence of ceratopsian dinosaurs and invertebrates of Lance age in the Medicine Bow formation, and Triceratops dinosaurs in the basal beds of the over- lying Ferris formation.^ The combined geologic data, exclusive of the plant evi- dence, are, therefore, consistent throughout in supporting a reference of the Medicine Bow formation as of Lance age, or more specificaUy as of HeU Creek age. The " Lovering, T. S., Aurand, H. A., Lavington, C. S., and WiUon, J. H., Fox Hills Formation, Northeastern Colorado. Amer. Aesoc. Petroleum Geologists BuU., vol. 16, 702. 1932. « Conway, W. P.. Jr., Slraiigraphy and Paleonlology of the Medicine Bow Formation, Hanna Baain, Wyoming, Dept. Geology, Princeton University, senior thesis, 30-32, 1936. 2 Dobbin, C. E., Bowen, C. F., and Hoota, H. W., U. S. Geol. Survey BuU. 804, 25, 1929. FOX HILLS AND LOWER MEDICINE BOW 7 Lance age of the IVIedicine Bow formation has previously been discussed by Hares ^ and substantiated by Dobbin and Reeside.^ NORTHWESTERN COLORADO The plant remains from this region were obtained from the succession of sand- stones, shales, and coal beds occupying the same stratigraphic position above the Lewis shale as do the sedinients of the Medicine Bow formation in southcrn Wyo- ming. These sediments have hitherto^been referred to the Laramic or "Laramie" formation. In view of their closer geographic position to the type locahty of the Medicine Bow formation than to that of the Laramie formation, as now restricted, it seems more appropriate to refer them to the former formation. Such a reference is furthcr substantiated by the fact that the outcrop of these strata can be traccd with only shght breaks ahi^ost continuously from the area in northwestern Colorado to the western part of the Hanna basin in southern Wyoming. Attention in this region was centered about the area surrounding Craig, Colorado, because of the previous reports of plant fossils in this region and because of their occurrence in sedimcnts conformably overlying beds which have yielded a Fox Hills fauna, including the index ammonite Sphenodiscus ^ which has been used throughout my work as a reference zone. The first important contribution to the geology of this region was the report on the Yampa coal field by Fenneman and Gale.^ The sediments conformably overlying the Lewis shale were here referred to the Laramie formation on the basis of stratigraphic position. A more comprehensive report by Gale on the coal fields of northwestern Colorado and northeastern Utah was pubhshed in 1910.^ The name Laramie formation was still retained for the strata above the Lewis shale on the basis of a few species of invertebrates and plants. In 1925, Hancock, ^ who was followed by Sears,' expressed doubt regarding the reference of these same beds to the Laramie formation by placing the name in quotation marks. Additional marine invertebrates, including the index species Sphenodiscus lenticularis (?), were ob- tained by Hancock from the lower massive sandstones of his "Laramie" formation, indicating the Fox Hills age of these beds. In the sections wliich we measured a few miles west of Craig, additional Fox Hills species were coUected from the massive sandstones above the Lewis shale. The same criteria for the recognition of the Fox Hills formation as were appUed in southern Wyoming (see p. 6) were found to obtain in this region. AU of the plant remains secured from this region were collected from horizons in the lower Medicine Bow formation, lying within a few hundred feet above the massive sandstones here referred to the Fox Hills formation. No attempt was made to determine the upper limit of the Medicine Bow formation, although it clearly ' Harea. C. J., Waehington Acad. Sci. Jour., vol. 5, 328-330. 1915. 2 Dobbin, C. E., and Reeeide, J. B., Jr.. U. S. Geol. Survey Prof. Paper 15S-B. 23, 1929. 'Hancock, E. T., U. S. Geol. Survey BuU. 757, 22, 1925; Dobbin. C. E., and Reeside, J. B.. Jr., op.cit., 23. « Fenneman, N. M., and Gale, H. S., U. S. Geol. Survey BuU. 297, 1906. ' Gale, H. S., U. S. Geol. Survey BuU. 415, 1910. « Hancock, E. T., op. cit., 21. 'Sears, J. D., U. S. Geol. Survey BuU. 751, 269-326, 1925. 8 C O N T R I B U TI O N S TO PAL^ONTOLOGY includes all but the basal 475 feet of what has hitherto been termed the "Laramie" formation. The "Laramie" has previously been separated by a persistent con- glomerate from the overlying Post-" Laramie " (Fort Union?) formation. The vaUdity of the Post-" Laramie " formation as a separate Mthologic unit has as yet not been fuUy corroborated by fossil evidence. The changes in the interpretation of the stratigraphic sequence in this region are shown in Figure 3. The formations below the Mesaverde group and above the Wasatch formation have been omitted as irrelevant to the present problem. Fenneman an(i Gale, 1907 Gale, 1910 Series Formation Thicknesa (feet) System Formation Thicknesa (feet) Eocene? (not designated) — Tertiary Wasatch 3400 Cretaceous Laramie 900+ Post-Laramie (Fort Union?) 800+ Cretaceous Laramie 1200 Lewis 1200+ Lewis 1200+ Mesaverde 3500 + Mesaverde 5280+ Sears, 1925; Hancock, 1925 * Present Interpretation Series Group Formation Thickness (feet) Series Group Formation Thicknesa (feet) Eocene Wasatch -Unconformity- Post-"Laramie" -Unconformity- "Laramie" 3000- 5600 0-800 1020- 2350? Eocene Wasatch -Unconformity- Post-"Laramie" -Unconformity? Medicine Bow Not Determined Not Determined 1200+ Eocene? Paleocene? Upper Cretaceous Upper Cretaceous Fox Hills 475 + Lewis 900- 1600 Lewis 1600 Mesaverde Williams Fork 1600- 3400 Mesaverde Williams Fork 1600 Iles 1350 Iles 1350 * Hancock'B section did not include the Poat-" Laramie " and Wasatch formationa. FiQ. 3 — Late Cretaceous and early Tertiary formations in northwestern Colorado. F O X H I L L S A N D L O W E R M E D I C I N E B O W 9 LOCAL SECTIONS AND MAPS CORSON RaNCH, WyOMING, LoCAUTIES P. 371 AND P. 372 Along the North Platte River in sections 2 and 3, T. 22 N., R. 85 W., thc valley wall exposes a sectioii from thc upper part of the L(>wis shale through the Fox Hills for- mation and into the ]Medicine Bow formation. The leaf horizons in the Mcdicine Bow formation are located about a half mile north of the line of the section, as foUows: Locality P. 371: SE. \ sec. 34, T. 23 N., R. 85 W., 275 yards N. 75° W. of the southeast section corner of sec. 34. Localily P. 372: SW. \ sec. 35, T. 23 N., R. 85 W., 260 yards N. 53° W. of the half-mile marker of the lower side of sec. 35. The exact positions of these locahties and the areal distribution of formations in the region are shown in figure 4. The measured section, indicating Uthology, R.es w R. 84 W, R. 83 w. SEDIMENTARY ROCKS Alluvium ■J S Fault -^— Axis of anticline FiG. 4— Geological setting of Localities P. 371, P. 372, and P. 373. After U. S. Geological Survey Bull. 804, pl. 27. 10 C^ONTRIBUTIONS TO PAL^ONTOLOGY <3§ go. 50; FOX HILLS AND LOWER MEDICINE BOW 11 thickness, and the position of faunal and floral horizons, as shown in figure 5, is as foUows: Sedion in the Lewis, Fox Hills, and Medicine Bow formations in secs. 2 and 3, T. 22 N., R. 85 W., Wyommg ' Thtckness Medicine Bow f ormation (lower part) : ^ " -^^^' Sandstone gray, soft, massive, witli concretionary lenses alternating with lignitic shales Hard ycllowish 4-foot layers of sandstone near middle and top . . 7^ Sandstone grav, hard, thinlv bedded, weathering to light brown. Hard gray bed near base yielded the following species of fossil plants (Locahty P. 3/2): Bracts (?) Myrica torreyi Canna cf C? magnifolia Palmocarpon? subcyhndricura Carpites gluma;formis Pandanites corsoni Celastrus taurinensis Phyllites colubrmoides Cinnamomum afRne Phyllites sp. Cissites lobatus Pistacia eriensis Cyperacites sp. Quercus viburmfoha Dombeyopsis obtusa Rhamnus cleburm Dombevopsis triviahs Rhamnus sahcifohus Dryophyllum subfalcatum Sabalites eocemca Equisetum sp. Sabalites montana Ficus cockereUi Sequoia? acumuiata Ficus crossii Trochodendroides nebrascensis Ficus planicostata Typha sp. Ficus trinervis Viburnum marginatum Grewiopsis saportana Viburnum montanum Leguminosites arachioides minor Zingiberites dubius Magnoliophyllum cordatum Zizyphus hendersom This sarae bed contained the fresh-water raolluscs: Unio danai Meek and Hayden Goniobasis tenuicarinata Meek and Hayden • ^" Sandstone, grayish-white, soft, massive; basal beds with the fresh-water moUuscs: Campeloraa nebrascensis nebrascensis Meek and Hayden Goniobasis tenuicarinata Meek and Hayden Proparreysia holmesiana White Tulotoma thompsoni White Unio baueri Stanton U. priscus Meek and Hayden n. var.? ^° Coal bed • • ■ ,••••••.•• u ' j j ' j Shale, gray, soft, intercalated with thin lignitic shales and massive to cross-bedded soft, gray sandstones. Horizon 30 feet frora base yielded petnfied wood and a single brackish-water raollusc : Corbula subtrigonalis perundata Meek and Hayden .' ou Sandstone, soft, gray, cross-bedded, raassive ■ ^ Shale, gray to brown, sandy, and soft, gray raassive sandstone »o Shale, gray and yellowish-brown with a few thin, massive, soft, gray sandstones chiefly covered in valley •.• • • .• • • • • • • • ,■ Shale, gray to brown, with 2 thin lignite layers and thm brecciated yellow sand- stone layers at top . „ Sandstone, grayish-white, hard, raassive • • ■ • •. Shale, gray to brown, with intercalated thin Ugmtic shales and massive gray Siiricl^toriGS Sandstone, gray, hard, raassive with thin brecciated bed at base 20 Shale gray to brown, gypsiferous, with a few thin, hard gray sandstone lenses. One of these lenses, 15 feet from top, j-ielded the foUowmg species of fossil plants (Locality P. 371): > Nomenclature of invertebrate genera and speciea foUows Henderson, J., Geol. Soc. Am., Spec. Pap. No. 3. 1935. 12 CONTRIBUTIONS TO PAL^ONTOLOGY Aneimia sp. Ficus trinervis Asplenium? coloradense ?JugIans praerugosa Celastrus taurinensis ?Magnolia dakotana Cyperacites sp. Magnolia pulchra Dryophyllum subfalcatum Magnohophyllum cordatum Dryopteris? carbonensis Phylhtes cokibrinoides Equisetum sp. PhyOites conwayi Ficus cockereUi Pistacia eriensis Ficus planicostata Rhamnus sahcifohus Ficus tesselata Sabahtes eocenica 54 Shale, gray, soft, yehowish-brown, gyiDsiferous, with massive, soft, gray sandstone lenses and three 1-foot coal beds. Horizon near middle with shells of the fresh- water mohusc : Unio danai Meek and Hayden 216 Sandstone, hard, gray, massive, concretionary at top 20 Sandstone, gray, soft, nodular, with 4- to 6-inch gray shale layers 96 Sandstone, yellowish-brown, hard, and intercalated brown and gray shales; few thin carbonaceous layers. Horizon near middle with shells of brackish-water mollusc : Ostrea glabra glabra Meek and Hayden 30 Shale, gray to yellowish-brown, with massive, gray sandstone and concretionary gray sandstone at top, containing shells of the brackish-water mollusc : Ostrea sp 66 Sandstone, gray, massive, lenticular and concretionary near top 12 Shale, yellowish and gray, with 3-foot coal bed in middle 15 Sandstone, gray, massive, and concretionary •_ • • 17 Shale, yellowish-brown, with intercalated layers of gray shale, thin lenses of brownish hard sandstone, and 3 thin coal beds 128 Shale, gray, sandy 15 Coal bed 3 Fox Hills f ormation : Sandstones, gray, concretionary, intercalated with gray, sandy shales 267 Sandstone, grayish-white, massive, with intercalated 6-inch layers of gray shale. Horizon near middle contains numerous shells of the foUowing marine moUuscs: Baculites sp. Cardium whitei Dall Corbula subtrigonalis perundata Meek and Hayden C. subtrigonalis subtrigonalis Meek and Hayden Corbula sp. Dentahum sp. Discoscaphites sp. Gervillia recta Meek and Hayden Melania insculpta Meek M. wyomingensis Meek Pteria (Pseudoptera) fibrosa Meek and Hayden Tancredia americana Meek and Hayden Tellina equilateralis Meek and Hayden Yoldia scitula Meek and Hayden 83 Shale, grayish-white, sandy 20 Sandstone, gray, hard, concretionary _2£ Total Fox Hihs formation 385 Lewis shale: Shale, grayish-white, sandy near top, grading down into darker gray shales. Thick- ness not measured. FivE MiLEs NoRTH OF Walcott, Wyoming, Locality P. 373 In this region a fairly continuous section from the Lewis shale through the Fox Hills formation and into the Medicine Bow formation is exposed about a mile FOX HILLS AND LOWER MEDICINE BOW 13 east of thc Walcott-Ferris ranch road, 5 miles N. 15° E. of the town of Walcott. Collections of both invertebrates and plants were secured from several horizons. The exact position of LocaUty P. 373, from which the plant rcmains werc obtained, is in the SE. \ scc. 36, T. 22 N., R. 84 W., shown in figurc 4. The dotails of the mcasured section and the position of faunal and floral horizons (figure 5) are as follows : Sedion in the Lewis, Fox Hills, and Medicine Bow formations in sec. 1, T. 21 N., R. 84 W., and sec. 36, T. 22 N., K. 84 W., Wyo7ning Thickness Medicine Bow formation (lower part) : ^" f^^^ Shale, gray, with thinly bedded, gray sandstone lenses 2-5 feet thick at irregular intervals. Horizon 20 feet from top contains fresh-water molluscs: Corbicula sp. Goniobasis tenuicarinata Meek and Hayden Horizon 29 feet from top yielded the following species of fossil plants: Carpites walcotti Pistacia eriensis Cinnamomum affine Rhamnus saUcifoHus CjTeracites sp. Sabalites eocenica Dombeyopsis trivialis Trochodendroides nebrascensis Ficus planicostata Viburnum marginatum Leguminosites arachioides minor Zizyphus hendersoni Myrica torreyi • ■ • • ; Shale, gray to brownish, sandy, with occasional thin irregular lenses of concre- tionary grayish-brown sandstones. Fresh-water molluscs at top: Campeloma nebrascensis nebrascensis Meek and Hayden Tulotoma thompsoni White Unio sp. Leaf horizon at base: Carpites walcotti Pistacia eriensis Celastrus taurinensis Rhamnus cleburni Dryophyilum subfalcatum Rhamnus sahcifoUus Ficus planicostata Trochodendroides nebrascensis Magnohophyllum cordatum Viburnum marginatum Myrica torreyi 20 Shale, gray, sandy, with few thin lignitic shale beds and intercalated thin gray sandstones. Fresh-water molluscs one foot from top: Campeloma nebrascensis nebrascensis Meek and Hayden Unio priscus abbreviatus Stanton and Hatcher U. priscus n. var.? Fossil leaves at horizon 20 feet from top: Carpites walcotti Pisonia? racemosa Celastrus taurinensis Pistacia eriensis Cinnamomum affine Quercus viburnifoHa Dombeyopsis trivialis Rhamnus cleburni Ficus planicostata Rhamnus salicifolius MagnoHophyllum cordatum SabaHtes eocenica Myrica torreyi Trochodendroides nebrascensis Palffioaster inquirenda Viburnum marginatum 51 Shale, gray to brown, intercalated with thin lenses of concretionary sandstone, two Hgnitic shale beds near base "*^ 14 CONTBIBUTIONS TO PAL^ONTOLOGY Shale, gray, and thinly bedded gray sandstone. Following species of plants col- lected at base: Carpites glumseformis Magnoliophyllum cordatum Carpites walcotti Myrica torreyi Dombeyopsis trivialis Pistacia eriensis Dryophyllum subfalcatum Quercus viburnifoHa Ficus cockerelli Rhamnus cleburni Ficus planicostata Rhamnus sahcifoHus Grewiopsis saportana Sequoia? acuminata Juglans leconteana Viburnum marginatum Magnoha dakotana Zizyphus hendersoni 61 Shale, dark gray, carbonaceous, thin lignites and gray massive to regularly bedded sandstone near base 48 Shale, gray to brown, sandy, with massive ripple-marked and concretionary gray sandstone at top 77 Shale, gray, sandy, grading upward into thinly bedded, gray sandstone and single coal bed at top. Fossil leaves at base: Cinnamomum UnifoUum Rhamnus cleburni Ficus planicostata Rhamnus sahcifolius Myrica torreyi Trochodendroides nebrascensis .... 38 Sandstone, gray, regularly thinly bedded 6 Shale, gray, sandy, with few exposed thin beds of gray, concretionary sandstones; hgnite fioat ; mostly covered 216 Sandstone, grayish-brown, concretionary, with few intercalated shale beds 31 Shale, soft; mostly covered 30 Shale, gray, sandy, and intercalated, thin to massively bedded gray sandstones, 2-f oot coal bed near base 97 Covered in valley, probably shale 60 Sandstone, gray, persistent bed, somewhat concretionary 11 Shale, dark gray, carbonaceous 18 Sandstone, gray, thinly bedded 35 Shale, gray to brownish-gray, sandy 20 Sandstone, gray, soft, thinly bedded 10 Shale, gray to brownish-gray, sandy 40 Coal bed 4 Shale, dark gray, soft 10 Coal bed 2 Sandstone, gray, massive 10 Sandstone, gray, soft, thinly bedded, with concretionary lens near middle. Frag- ments of brackish-water mollusc : Ostrea sp 38 Sandstone, gray, hard, lenticular, concretionary 6 Shale, gray to brownish-gray, sandy, with few thinly bedded, soft gray sandstone layers and intercalated lenses of gray, concretionary sandstones 88 Shale, dark gray, with 1-foot carbonaceous shales at base and top 15 Shale, dark gray, medium-bedded; thin concretionary layer at top yielded brackish- water molluscs: Melania wyomingensis Meek Ostrea glabra glabra Meek and Hayden Ostrea sp. Unio danai Meek and Hayden 10 Shale, gray and brownish, sandy 13 Sandstone, gray, massive, soft, with thin concretionary layer at top 27 Shale, gray, carbonaceous, to sandy near base, few poorly exposed beds of gray, concretionary sandstone, in large part covered 69 FOX HILLS AND LOWER MEDICINE BOW 15 Coal bed ^ Shale, gray, sandy ^ Fox Hills formation : Sandstone, gray, liard, iioorly bcddcd 6 Shale, soft, sandy, mostly covered 44 Sandstone, grayish-brown, poorly bedded to massive, with irregular, short lenses of gray concretionary sandstones. Horizon at top with numerous shells of the brackish-water species : Ostrea sp. Horizon at basc with numerous shells of marine spccies: Anatina sulcatina Shumard Melania wyomingensis Mcek Anatina? n. sp. Modiola attenuata Meek and Cardium whitei Dall Hayden DentaUum sp. M. herseyi White Discoscapliitcs cheyennensis Ostrea sp. (Owen) Mcek Pholadomya subventricosa Fasciolaria? cheyennensis Meek Sphairiola sp. and Hayden Sphenodiscus lenticularis (Owen) Legumen planulatum Conrad Meek Mactra nitidula Meek and TeUina scitula Meek and Hayden Hayden Tellina sp. M. alta Meek and Hayden Yoldia scitula Meek and Hayden ... 23 Shale, soft, gray, sandy, mostly covered in linear valley 112 Sandstone, soft, brownish-gray, with concretionary masses; wood fragments scattered 48 Shale, soft, grayish-brown, sandy, mostly covered 24 Sandstone, gray to brown, massive, cross-bedded, with few layers of grayish-brown sandy shale, f orms prominent scarp 125 Total Fox Hills formation 382 Lewis shale : Shale, gray, with minor amount of sand at top. Thickness not measured. Elk Mountain Road, Wyoming, Locality P. 374 The stratigraphic section at this locahty chffers only in details froni that of the Corson Ranch and Walcott localities. The exact position of Locahty P. 374 (NE. I sec. 13, T. 21 N., R. 81 W.) and the areal distribution of formations in this region are shown in figure 6. The occurrence of fossil leaves in the lower Medicine Bow formation had not been previously reported in this vicinity. The presence of the Sphenodiscus zone ' of Fox Hills age directed our attention to the region, resulting in the discovery of several leaf horizons in the overlying 1200 feet of beds. Additional coUections of both Fox Hills and Medicine Bow invertebrates were also secured. The measured section, indicating the Uthologic characters and the position of faunal and fioral zones (figure 5), is as foUows: I Dobbin, C. E., Bowen, C. F., and Hoots, H. W., U. S. Geol. Survey BuU. S04, 21, 1929. 16 CONTRIBUTIONS TO PAL^ONTOLOGY EXPLANATION FiG. 6— Geological setting of Locality P. 374. After U. S. Geol. Survey Bull. 804, pl. 27. Section in the Lewis, Fox Hills, and Medicine Bow formations in the NE. j sec. 13, T. 21 N., R. 81 W., Wyo. Thickness Medicine Bow f ormation (lower part) : in feet Sandstone, fine-grained, yellow-brown, massive 35 Sandstone, gray, 2- to 4-foot beds and interbedded dark gray shales 64 Sandstone, fine-grained, yellow-brown, massive 10 Shale, soft, sandy; mostly covered 155 Shale, gray, sandy, carbonaceous, and interbedded thin, gray, fine-grained sand- stones 28 Shale, soft, sandy, thinly bedded, gray to brown, interbedded with thinly bedded gray, quartzitic sandstones 48 Sandstone, massive gray-brown, fine-grained, cross-bedded near middle and at top, concretionary in upper third. Horizon 5 feet from base yielded the following fresh-water moUuscs : FOX HILLS AND LOWER MEDICINE BOW 17 Campeloma nebrasccnsis nebrascensis Meek and Hayden Goniabasis tenuicarinata Meek and Hayden Proparreysia holmcsianus White Tulotoma thompsoni White Unio cf. U. mendax White Unio sp 32 Sandstone, massive, fine-grained, friable, yellow-brown 23 Shale, soft, brown, papery, somewhat sandy 12 Sandstone, fine-grained, hght reddish-brown. In the basal portion were found the fossil plant species: Celastrus taurinensis Quercius viburnifoHa CornophyUum wardii Rhamnus cleburni Ficus cockerelli Rhamnus sahcifolius Ficus crossii SabaUtes montana Ficus planicostata Trochodendroides nebrascensis Grewiopsis saportana Typha sp. Laurophyllum mecki Viburnum marginatum Myrica torreyi 8 Shale, soft, sandy, gray, thinly bedded, in places carbonaceous 51 Shale, grayish-red, papery, with 2-foot coal bed at base, overlain by irregularly bedded, fine-grained, gray-brown sandstone 75 Shale, gray, papery, with thin, concretionary, fine-grained, fractured, gray-brown sandstone near middle 48 Sandstone, reddish-brown, cross-bedded, fine-grained 3 Sandstone, hght brown, fine-grained with shells of the brackish-water species: Corbula subtrigonaUs subtrigonaUs Meek and Hayden 15 Shale, sandy, partly carbonaceous with irregularly bedded, fine-grained, brown sandstone in middle part 33 Sandstone, brown, massive, friable, fine-grained 5 Shale, gray, sandy, papery ; 2-foot coal bed overlain by a single 4-foot layer of thinly bedded brown sandstone near middle 36 Sandstone, 2-foot layer, massive, brown, fine-grained, friable 2 Sandstone, gray, hard, thinly bedded, concretionary, which yielded the foUowing species of fossil plants: Apeibopsis? laramiensis MagnoUa pulchra Carpites walcotti MagnoUophyUum cordatum Cinnamomum affine Myrica torreyi CornophyUum wardii PhylUtes wilderi DryophyUum subfalcatum Quercus viburnifoUa Ficus coloradense Rhamnus cleburni Ficus cowanensis Rhamnus salicifoUus Ficus planicostata SabaUtes montana Ficus tesselata Trochodendroides nebrascensis Grewiopsis saportana Typha sp. Juglans leconteana Viburnum marginatum Juglans newberryi Viburnum montanum Laurus sociaUs? Vitis stantoni MagnoUa lakesii Zizyphus hcndersoni MagnoUa nervosa 5 Sandstone, gray, cross-bedded, concretionary, and lesser thickness of shales, gray, sandy 15 Sandstone, gray, massive, fine-grained " Shale, sandy, thinly-bedded; mostly covered 105 18 CONTRIBUTIONS TO PAL^ONTOLOGY Sandstone, gray, fine-grained, cross-bedded 4 Shale, sandy, gray ; mostly covered 53 Sandstone, gray, cross-bedded, fine-grained, quartzitic, with the f ollowing brackish- water species of molluscs : Anodonta? sp. Corbula subtrigonalis perundata Meek and Hayden C. subtrigonalis subtrigonalis Meek and Hayden 3 Shale, sandy; mostly covered 16 Coal bed 1 Shale, grayish-brown, papery 12 Sandstone, grayish-brown, thinly bedded, friable 5 Sandstone, gray, quartzitic, resistant, thinly bedded 3 Shales, gray, sandy, containing gypsum and numerous shells of the brackish-water ■ form: Ostrea sp 12 Sandstone, gray, quartzitic, fine-grained, thinly bedded, the bedding surfaces ripple- marked 5 Shale, grayish-brown, sandy 50 Shale, gray, papery, with shells of the brackish-water form: Ostrea sp 20 Sandstone, grayish-brown, fine-grained, crqss-bedded 10 Sandstone, gray, fine-grained, quartzitic 6 Shale, gray, sandy, with shells of the brackish-water form: Ostrea sp 10 Sandstone, grayish-brown, cross-bedded, friable 3 Shale, gray, sancly, partly covered 40 Coal bed, poorly exposed 2 Shale, gray, sandy, mostly covered 14 Fox Hills formation: Sandstone, grayish-brown, fine-grained, friable, massive, with 2-foot beds of thinly cross-bedded, gray sandstone in middle and at base 14 Shale, gray, sandy, and irregular lenses of gray, soft sandstone 20 Sandstone, gray, lenticular, resistant, thinly cros.s-bedded 2 Sandstone, soft, shaly, gray, interbedded in thin irregular layers with darker grayish- brown sandy shales, in part covered 178 Sand.stone, gray, massive, irregularly bedded, with large concretionary masses of harder, almost quartzitic, gray sandstone, intcrbedded with some gray, sandy shale. Horizon 6 feet from top yielded the foUowing marine species: Cardium whitei Dall M. alta Meek and Hayden Legumen planulatum Conrad Sphenodiscus lenticularis (Owen) Mactra warrenana Aleek and Meek Hayden TeUina scitula Meek and Hay- M. nitidula Meek and Hayden den 76 Sandstone, soft, gray, and shales, gray, sandy, mostly covered in valley. Thick- ness not measured. Craig, Coloeado, Localities P. 375 and P. 376 In this vicinity, dry gullies draining into the Yampa valley expose a discon- tinuous section from the massive sandstones here included in the Fox Hills forma- tion up into the lower beds of the Medicine Bow formation. The leaf-bearing horizons in the basal beds of the latter formation are located as follows: Locality P. 375: dry gully 200 yards west of the center of sec. 32, T. 7 N., R. 90 W., i mile W. 18° N. of the old Kimberly coal mine. F O X H I L L S A N D L O W E R M E D I C I N E B O W 19 Locality P. 376: southwest sidc of ridgo, SE. \ sec. 1, T. 6 N., R. 90 W., \ mile northcast of thc Kolasch raiich house. Both of these locahties were reported to mc by Dr. T. W. Stanton of the U. S. Geological Survey ^ as ha\ing yielded fossil plants from the basal beds of the so- called "Laramie" formation. The exact position of the localities and the arcal distribution of formations in this region are shown in figure 7. Alluviura SEDIMENTARY ROCKS , b ■z b: < C M Z ^^ >o o Cd Medicine Bow formation Fox Hills formation Lewis stiale Williams Fork formation IGNEOUS ROCKS < H a o a M Basalt FiG. 7— Geological setting of Localities P. 375 and P. 376. After U. S. Geological Survey Bull. 297. In view of the poor exposures in the vicinity of Craig, the measured section given below was taken farther west along the strike, in the dry valley walls of Spring Creek, 24 miles W. 10° N. of Craig. Here the section is well exposed in steeply dipping beds from the Williams Fork formation up through the Lewis, Fox Hills, Medicine Bow, and Post-" Laramie " formations. Exact stratigraphic positions with relation to the units exposed near Craig wereea sily determined by the persistence of the two massive, fossiliferous sandstones of the Fox Hills above the well-defined Lewis shales,^ and the presence in both regions of a persistent coal • Stanton, T. W.. written communication. June 8, 19.35. «Hancock, E. T., U. S. Geol. Survey BuU. 757, 22, 1925. 20 CONTRIBUTIONS TO PAL^ONTOLOGY bed lying between the massive sandstones of the Fox Hills. The precise position of the leaf-bearing horizons may be up to 100 feet in error owing to the shght thick- ening of the recognizable units toward the west. The section, showing lithology, thickness, and the position of faunal and floral horizons (see figure 5), is as follows: Section in the Lewis, Fox Hills, and Medicine Bow formations in sec. 10, R. 95 W., T. 7 N., Colorado Thickness Medieine Bow formation (lower part) : in feet Sandstone, grayish-brown, massive, resistant, lenticular, interbedded with thin, gray, papery, clay shales; shale bed at top contains the brackish-water form: Ostrea sp . 90 Coal bed 2 Shale, gray, sandy, with lenses of soft, gray sandstone 10 Sandstone, grayish-white, massive, resistant 10 Sandstone, gray, massive, interbedded with softer, thinly bedded, shaly sandstones 90 Shale, dark gray, sandy, with few interbedded lenses of soft sandstone 83 Sandstone, gray, massive, resistant; approximate position of resistant sandstone in the Craig area which yielded the following species of fossil plants (Locality P. 375): Carpites walcotti PhylUtes craigensis Cinnamomum affine PhyUites triUioides Credneria protophyUoides Quercus viburnifoUa Cyperacites sp. SabaUtes eocenica Ficus crossii Sequoia nordenskioldi Ficus planicostata Trochodendroides nebrascensis MagnoUophyUum cordatum Typha sp. Myrica torreyi " 3 Sandstone, grayish-white, soft, shaly, thinly bedded 42 Shale, soft, sandy, with thin lenses of fine-grained, quartzitic gray sandstones; approximate position of Kolasch ranch horizon which yielded the foUowing species of fossil plants (LocaUty P. 376) : Cinnamomum afRne Laurophyllum meeki Cinnamomum linifolium Quercus viburnifolia Cissites lobatus Rhamnus cleburni Cornophyllum wardii Rhamnus salicifolius Credneria protophylloides ?SabaUtes montana Dombeyopsis trivialis Sequoia? acuminata Dryophyllum subfalcatum Sequoia nordenskioldi Ficus crossii Trochodendroides nebrascensis Ficus planicostata Viburnum marginatum Ficus trinervis Vitis stantoni Juglans priErugosa 30 Sandstone, gray, shaly, soft, grading down into massive, cross-bedded sandstone . . 34 Fox Hills f ormation : Sandstone, grayish-white, fairly resistant, irregularly cross-bedded, with many scattered Umonite concretions; horizon 5 feet from top yielded the following marine invertebrates : Acteon attenuatus Meek and Hayden Cardium whitei Dall Dosiniopsis sp. Legumen planulatum Conrad Mactra nitidula Meek and Hayden Modiola herseyi White FOX HILLS AND LOWER MEDICINE BOW 21 Horizon 2 foet from basp yieldod marine invertebrates: Cardium whitei Dall Lunatia concinna Meek and Haydcn Lunatia dakotensis Henderson Mactra sp 167 Shale, gray, papery, and tiiinly bedded, soft sandstone; brackish-water species at base: Corbicula fracta fracta (Meek) Ostrea glabra arcuatilis Meek Ostrea sp 10 Sandstone, gray, massive, resistant, coarse; shelis of brackish-water form near middle : Ostrea sp 18 Shale, gray , sandy 3 Coal bed 4 Shale, gray, sandy 12 Sandstone, grayish-brown, concretionary, fine-bedding. Hancock reported the following brackish-water and marine species from this "lower massive sand- stone":' Anomia micronema Meek Cardium speciosum Meek and Hayden Ostrea glabra Meek and Hayden Scaphites? sp. Sphenodiscus lenticularis (Owen)? Tellina scitula Meek and Hayden 110 Shale, gray, sandy, with irregular lenses of grayish-brown, massive, fine-grained sandstones; horizon at top yielded brackish-water species: Anomia micronema Meek Ostrea glabra glabra Meek and Hayden Ostrea sp 52 Shale, gray, sandy, interbedded with three 1-foot layers of fine-grained quartzitic, gray sandstone; marine species at base: Acteon attenuatus Meek and Mactra warrenana Meek and Hay- Hayden den Anomia sp. Pholadomya sp. Baculites sp. Pteria sp. Cardium whitei Dall TelHna scitula Meek and Hayden DentaUura sp. Thetis circularis Meek and Hayden Legumen planulatum Conrad Yoldia evansi Meek and Hayden 26 Sandstone, grayish-brown, massive, fine-grained, resistant, with scattered remains of Halymenites 24 Shale, gray, papery, grachng down into more sandy, thinly bedded shales; horizon 5 f eet f rom top yielded f resh-water species : Campeloma nebrascensis nebrascensis Meek and Hayden Unio cf. U. priscus priscus Meek and Hayden Unio sp 35 Sandstone, gray, concretionary, resistant, with 1-foot lenses of finer-grained, gray sandstone at base containing marine species: Anatina sulcatina Shumard Dosiniopsis sp. Anchura sp. Micrabacia americana Meek and BacuUtes sp. Hayden Cardium sp. Tellina scitula Meek and Hayden _20 Total Fox Hills formation 481 Lewis shale: Gray, soft, clay shales, in part papery, in part sandy. Thickness not measured. ' Hancock, E. T., U. S. Geol. Survey BuU. 757, 22, 1925. COMPOSITION OF THE FLORA SYSTEMATIC RELATIONSHIPS The flora of the lower Medicine Bow formation, as here described, comprises 64 recognizable forms. Of these, 47 are referred to species previously described by others, 11 are described as new species, and 6 are recorded as not specifically deter- minable. The various divisions of the plant kingdom are represented as follows: Dicotyledonse, 45 species; Monocotyledonse, 8 species; Coniferales, 2 species; Equisetales, 1 species; Filicales, 3 species; position uncertain, 5 species. Within most of these divisions, and in particular within the Dicotyledonae, the systematic allocation to existing genera, and in some cases to famihes, is subject to revision. As is discussed under each species in the section on Systematic Palseobotany, refer- ences to modern genera have previously been made on the basis of leaf characters which in many cascs seem only remotely comparable to the designated genera. Such forms might better have been placed in non-committal form genera until con- clusive evidence of their systematic status had been obtained, but are here included under their assigned names to avoid confusion. In the following systematic list of species of the lower Medicine Bow flora, the genera marked with an asterisk are of doubtful botanical significance. Pteridophyta Filicales *Aspleni.umf coloradense Knowlt.on *Dryopterisf carbonensis Knowlton *Aneimia sp. Equisetales Equisetum sp. Spermatophyta Gymnospcrmse Coniferales Sequoia nordenskioldi Heer *Scquoiaf acuminata Lesquereux Angiospermse Monocotyledonse Typhaceae Typha sp. Pandanaceae Pandanites corsoni Dorf Cyperacese Cyperacites sp. Palmse Sabalites montana (Knowlton) Dorf Sabalites eocenica (Lesquereux) Dorf *Palmocarponf subcylindricum Lesque- reux CannaceEe Canna cf. C.f magnifolia Knowlton Zingiberacese *Zingiberites dubius Lesquereux Dicotyledonse Myricacese *Myrica torreyi Lesquereux 22 Juglandaceie *Juglans leconteana Lesquereux Juglans newberryi Knowlton Juglans prccrugosa Knowlton Fagacese Dryophyllum subfalcatum Lesquereux *Que7Xus viburnifolia Lesquereux Moracese Fictis planicostata Lesquereux *Ficus cockerelli Knowlton *Ficus trinervis Knowlton *Ficus crossii Ward *Ficus tesselata Lesquereux Ficus coloradensis Cockerell *Ficus cowanensis Knowlton Platanacese Credneria protophylloides Knowlton Nyctaginacese *Pisoniaf racemosa Lesquereux Lauracese Cinnamomum linifolium Knowlton Cinnamomum affi,ne Lesquercux Laxirophyllum meeki Dorf Laurus socialis Lesquereux? Trochodendrace se Trochodendroides nebrascensis (New- berry) Dorf Magnoliacese Magnolia lakesii Knowlton Magnolia dakotana Berry FOX UILLS AND LOWER MEDICINE BOVV 23 Magnolia 7iervosa (Knowlton) Brown Alagnolia pulchra Ward Magnoliophyllum cordatum Dorf Leguminosse Leguminosites arachioides minor Berry AnacardiacetE Pistacia eriensis Knowlton Celastraceaj *Celastrus taurinensis Ward Riiamnaeeae *Rhamnus salicifolius Losqucreux *Rham7i^is cleburni Lcsqucrcux Zizyphus hendersoni Knowlton Vitaccffi Cissites lobatu^ Dorf Vitis slantoni (Ivnowlton) Brown TiliacejB *Apeibopsis? laramiensis Knowlton Grewiopsis saportana Lesquereux Cornacea; Cornophyllum wardii Dorf Bignoniaceaj *Dombeyopsis obtusa Lesquereux *Dombeyopsis trivialis Lesquereux Caprifoliaceaj Viburrium marginatum Lesquereux Viburnum montanum Knowlton Position Unccrtain PalcEoaster inquirenda Knowlton Carpites walcotti Dorf Carpites glumcrformis Lesquereux Phyllitcs craigensis Dorf PhylUtes wilderi Dorf Phyllites conwayi Dorf Phyllites colubrinoides Dorf *Phyllites trillioides Dorf Phyllites sp. Bracts (?) QUANTITATIVE ANALYSIS For the purposes of both ecologic inferences and geologic correlation, it has become increasingly instructive to make quantitative leaf-counts at locahties where a large number of individual specimens can be obtained. In the Medicine Bow formation only 3 localities have furnished a sufficient number of specimens for consideration. At all 6 locaUties the plant-bearing horizons were found to be within discontinuous, lenticular units, making extensive quarrying for quantitative studies impossible. At those locahties where less than 200 specimens could be secured a careful record was kept of the order of dominance of the various species. The leaf counts at Locahties P. 372, P. 374, and P. 376 are shown in table 1. At the remaining 3 localities the order of relative abundance of those species rep- resented by more than 5 specimens is as f ollows : Locality P. 371 1. Dryophyllum subfaleatum 2. Ficus triner^ns 3. Ficus tesselata 4. Pistacia eriensis 5. Ficus planicostata 6. Phyllites colubrinoides 10. Cinnamomum affine The most abundant leaves are clearly those of Ficus planicostata, with 25.6 per cent of the 1700 specimens from 3 localities, and 5th, Ist, and 3rd, respectivelj', Ln order of dominance at the other 3 localities. The remaining dominants in their apparent relative order are as follows: Viburmmi marginatum, Myrica torreyi, Dryophyllum subfalcatum, Grewiopsis saportana, Magnoliophyllum cordatum, Rham- nus salicifolius, Rhamnus cleburni, and Trochodendroides nebrascensis. These 9 Locality P. 373 Locality P. 375 1. Ficus planicostata 1. MjTica torreyi 2. Myrica torreyi 2. Sequoia nordensliioldi 3. Viburnum marginatum 3. Ficus planicostata 4. Rhamnus salicifolius 4. Magnoliophyllum cordatum 5. Rhamnus cleburni 5. Sabalites eocenica 6. Pistacia eriensis 6. Typha sp. 7. Carpites walcotti 7. Trochodendroides 8. MagnoHophyllum cordatum nebrascensis 9. Trochodendroides nebrasceusis 24 C O N T R I B U T I O N S TO PAL^ONTOLOGY species make up a total of 83.7 per cent of the 1700 specimens from Localities P. 372, P. 374, and P. 376. At the remaining 3 localities these same species are consistently among the more abundant. Table 1 — LeaJ counts at three localUies Species Locality P. 372 No. p.ct. Locality P. 374 No. p.ct. Locality P. 376 No. p.ct. Totals No. p.ct. 1 . Ficus planicostata 2. Viburnum marginatum 3. Grewiopsis saportana 4. Myrica torreyi 5. Dryophyllum subfalcatum . . . . 6. Magnoliophyllum cordatum . . 7. Rhamnus cleburni 8. Rhamnus salicifolius 9. Trochodendroides nebrascensis 10. Dombeyopsis triviaiis 11. Cornophyllum wardii 12. Cissites lobatus 13. Cinnamomum affine 14. Quercus viburnifolia 1.5. Ficus crossii 16. Credneria protophylloides .... Totals Total number of leaves counted . . . 390 241 3 16 43 81 67 11 2 20 21 14 7 12 39.0 24.1 0.3 1.6 4.3 8.1 6.7 1.1 0.2 2.0 2.1 1.4 0.7 1.2 29 4 170 151 37 2 4 24 31 2 20 3 2 2 5.8 0.8 34.0 30.2 7.4 0.4 0.8 4.8 6.2 0.4 4.0 0.6 0.4 0.4 16 19 36 11 31 4 9 8 6 4 10 3 10 8.0 9.5 18.0 5.5 15.5 2.0 4.5 4.0 3.0 2.0 6.0 1.5 5.0 435 264 173 167 116 83 82 66 37 31 28 27 21 19 17 10 25.6 15.5 10.2 9.8 6.8 4.9 4.8 3.9 2.2 1.8 1.6 1.6 1.2 1.1 1.0 0.6 928 92.8 481 96.2 167 83.5 1576 92.6 1000 500 200 1700 Where fossil fiorules can be coUected from a number of scattered localities in a single formation, reUable information concerning relative abundance of species can also be derived from a consideration of those species which occur at the largest number of locaUties. The relative order of dominance from this point of view, summarized from the local distribution of species at the 6 Medicine Bow locaUties (table 4), is as foUows: 1. Ficus planicostata 6 localities 2. Dryophyllum subfalcatum 5 3. Magnoliophyllum cordatum 5 4. Rhamnus salicifolius 5 5. Trochodendroides nebrascensis 5 6. Quercus viburnifolia 5 7. Cinnamomum affine 5 8. Viburnum marginatum 4 9. Myrica torreyi 4 10. Rhamnus cleburni 4 A comparison of this Ust with the 9 dominants based on leaf counts shows that 8 of the latter are among the 10 most widely distributed at the various Medicine Bow localities. There is thus a considerable degree of correspondence between the most abundant species in the individual florules and the most widespread species in the flora as a whole. Only a single species, Grewiopsis saportana, among the FOX HILLS AND LOWER MEDICINE BOW 25 9 leaf-count dominants is recorded from lcss than 4 of thc 6 locahties. On the other hand therc are 2 species, Quercus viburnifoUa and Cinnamormim ajjine, which are more common in distribution than in individual abundance. LEAF CHARACTERS An analysis of certain structural characters of modern dicotyledonous lcaves of a given area can, with reasonable accuracy, be utihzed to indicate the environ- mental conditions of growth. Bailey and Sinnott ' have studied such leaf charac- ters as size, venation, and margin in fioras of the various chmatic zones of the world and have shown that therc is, in the first place, a gradual increase in the proportion of entire-Ieaved dicotj^ledons from the forests of the cold temperate zone southward into the warm temperate, subtropical, and tropical zones. A few of their repre- sentative areas are shown in the following list : Per cent entire-leaved Cold temperate zone: Central Russia 28 East-central North America 28 France 44 Warm temperate zoi\e: East-central China 48 Southeastern United States 49 Los Angeles region 54 Subtropical and tropical zone: Hongkong 71 Bombay 72 West Indies 76 Florida 83 Malay States 86 Sinnott and Bailey also pointed out that large leaves, compound leaves, and pin- nate leaves are most abundant in the lowland tropics and decrease proportionately into the subtropical and temperate zones.^ The application of several of these criteria to fossil floras was suggested and attempted by Bailey and Sinnott with considerable success. Among others, the early Tertiary Florissant and Green River floras of the Rocky Mountains, with 33 and 29 per cent, respectively, of entire-margined leaves appeared to indicate temperate conditions. In contrast, the Eocene Wilcox flora of the Gulf coast, with 83 per cent of entire-margined leaves, indicated subtropical conditions. These general conclusions have been amply corroborated by the studies of the genera and species of the several floras in question. Chaney and Sanborn ^ have more recently applied similar criteria to the Eocene Goshen flora of Oregon, and have added the textural character and the presence or absence of a dripping point as useful leaf characters for the determination of climatic conditions. Their conclusion that ' Bailey, E. W., and Sinnott, I. W., Science, n. s., vol. 41. 832, 1915. «Sinnott, I. W., and Bailey, E. W., .\mer. Jour. Bot., vol. 2, table I, 1915. ' Chaney, R. W., and Sanborn, E. I., Carnegie Inst. Wash. Pub. No. 439, 18, 1933. 26 C O N TR I B U T I O N S TO PAL^ONTOLOGY the Goshen flora was subtropical is in close agreement with their inferences based on the modern distribution of related species. The results of the analysis of leaf characters of the dicotyledons in the Medi- cine Bow floras are shown in table 2. Also introduced in this table are the analyses Table 2 — Leaf charactcrs of dicntyledms, showing relative percentages in each category Margin Length Nervation Texture Organization Dripping Point En- tire Non- entire Over 10 cm. Under 10 cm. Pin- nate Pal- mate Thick Thin Sim- ple Com- pound Pres- ent Ab- sent Muir Woods, 22 species. . . . Bridge Creek, 20 species Weaverville, 36 species Medicine Bow, 42 species . . . Laramie, 55 species Goshen, 49 species Panama forest, 41 species . . 23 15 47 67 71 61 88 77 85 53 33 29 39 12 27 30 60 45 45 53 56 73 70 40 55 55 47 44 77 70 60 60 64 82 83 23 30 40 40 36 18 17 64 55 57 71 73 98 98 36 45 43 29 27 2 2 77 80 81 83 87 88 85 23 20 19 17 13 12 15 9 10 49 48 18 47 76 91 90 51 52 82 53 24 of the Goshen flora, the Bridge Creek flora, the modern flora of the Muir Woods, Cahfornia, the modern Panama flora,^ the Eocene WeaverviUe flora of Cahfornia,^ and the Upper Cretaceous Laramie flora of Colorado. In compiling table 2, I have used a somewhat different arrangement from that of Chaney and Sanborn, and of MacGinitie. The characters of margin, length, and nervation are placed in the first three columns, as they are considered more reliable and more diagnostic than the remaining characters. There are two reasons for so considering them: (1) these characters have been tested more fuUy in modern forests, chiefly by Bailey and Sinnott; (2) they are more readily and more accu- rately determinable in fossil leaves than are the remaining characters. A quan- titative determination of texture in fossil leaves is somewhat more difficult, since it involves a considerable amount of personal opinion. Likewise in organization, it is often impossible to ascertain whether a fossil leaf is actually a simple leaf or a leaflet of a compound leaf. It is usually necessary in this category to resort to comparisons of the fossil leaves with modern correlatives, a procedure which com- pletely defeats the purpose of this study, namely to make climatic inferences which shall check and supplement the inferences made on the basis of taxonomy and modern distribution of correlatives. The presence or absence of dripping point has not yet been statistically studied in a sufficient number of modcrn forests to be rehably used. It has been my experience, moreover, that it is often a matter of opinion as to just how abrupt and elongate a leaf tip must be to be considered a dripping point. I have also been forced to consider dripping points absent from many fossil leaf species merely because the tip was not preserved, a condition which actually may have resulted from the very presence of a fragile, easily destroyed dripping point. ' Chaney and Sanborn, ibid., 19-21. 2 MacGinitie, H. D., Carnegie Inst. Waah. Pub. No. 465, 113, 1937. FOX HILLS AND LOWER MEDICINE BOW 27 In the interprctation and application of the data presentcd in table 2, it is evidcnt in the first placc that on the basis of margin charaeters the Aledicine Bow association is comparablc to forests approaching the subtropical in aspect. The large numbcr of entire-leaved spccies (67%) is only 4% less than Bailey and Sin- notfs lowest figure for sul)tropical forests. Among fossil fioras the Mediciiie Bow asscmblage is of decidedly warmcr aspect than the floras of the Bridge Creck (15%), the Green River (29%), and the Florissant (35%); the Wilcox flora, on the other hand, appears to be more subtropical (83%). The Goshen and Laramie floras are more nearly similar to the Medicine Bow in percentage of entirc-leaved species. The Goshen flora (61%), according to Chaney and Sanborn.i indicatcs a humid, subtropical climate. The Laramie flora (71%) has unfortunately not as yet been thoroughly studied with a view to accurate postulation of climatic conditions. In the column under " Length" in table 2, the percentage of large leaves in the Medicine Bow flora shows likewise its closest relationship to that of the Goshen flora. In nervation characters a closer similarity is shown to the Weaverville flora, which MacGinitie ^ has described as a warm temperate association, com- parable to the modern flora of southern Alabama. Although the remaining leaf cbaracters are not regarded as sufficiently trustworthy criteria for individual con- sideration, they tend to substantiate the similarity of the Medicine Bow species to those of the subtropical Goshcn flora on the onc hand and the warm temperate Weavervillc flora on the other. To summarize the inferences made on the basis of reliably tested leaf charac- ters, it is evident that the Medicine Bow flora is distinctly not a cool temperate assemblage, but is intermcdiate in character between warm temperate and sub- tropical, morc nearly approaching the latter. PAST CLIMATIC CONDITIONS AS INDICATED BY THE FLORA In addition to climatic studies based upon leaf characters, environmental con- ditions of the past may also be cstimated by comparing thc fossil species of a given flora with their nearest living relatives; it may be inferred that the past conditions were analogous to those under which the majority of living relatives grow. The application of this method to the lower Medicine Bow flora encounters several difiiculties. In the first place, many of the fossil species are not generically comparable to any living forms which have come to my attention. This may mean one of at least three tliings: (1) these genera may not have survived from the Upper Cretaceous to the present; (2) they may be ancestral types from which descendants with somewhat different leaf characters have since evolved; or (3) they may belong to genera with which I am unfamiliar. In the second place, even when identifications are more certain, it cannot be estabhshed that in the Upper Creta- ceous a gcnus responded to climatic influences precisely as it does today. This difficulty is in a large measure overcome by placing less stress upon individual genera in the fossil flora and more upon the flora as a whole. Thirdly, where close ' Chaney, R. W.. and Sanbom, E. I.. Carnegie Inst. Wash. Pub. No. 439, 57, 1933. 2 MacGinitie, H. D.. Carnegie Inst. Wash. Pub. No. 465, 113, 130, 1937. 28 CONTRIBUTIONS TO PAL^ONTOLOGY relationships have been estabhshed between fossil forms and hving genera and species, it is often difficult to determine from either Uterature or herbarium speci- mens the precise distribution and climatic requirements of the hving relatives. This appUes particularly to subtropical and tropical forms. To anyone familiar with temperate forests of either hemisphere, it is apparent that the lower Medicine Bow flora is distinctly not a cool temperate assemblage. The absence of t^qDicaUy cool temperate genera is striking. Rather than rely on such negative evidence, I would emphasize that the majority of the lower Medicine Bow species which are comparable to hving forms show close resemblances to genera now hving in warm temperate and subtropical forests, as shown in table 3. Fossil species which have not been found comparable to anj^ modern leaves have been omitted from this hst. The distribution of modern correlatives and alternatives shown in table 3 indicates that only a small number of genera are found in cool temperate assem- blages. With no exceptions, moreover, these few genera are of non-restricted distribution, extending in all cases into warm temperate or subtropical associations. Some of them, such as Nyssa, Magnolia, Drimys, and Viburnum are in fact more numerous and widespread in warmer forests, and extend into the cool temperate regions only in a Umited number or a Umited area. The large number of generic correlatives of warm temperate and subtropical forests clearly substantiates the inferences based upon the characters of leaves of the Medicine Bow species. The relatively large number of tropical genera represented does not imply close relation- ship to lowland tropical forests, since none of these genera is restricted to such habitats; aU of them are equaUy weU developed in subtropical associations, and may range into the warm temperate. A consideration of several individual famUies whose presence is weU estab- Ushed in the Medicine Bow flora adds weight to the cUmatic inferences based upon leaf characters and the aspect of the flora as a whole. Most convincing, perhaps, is the famUy Palmse, represented by Sabalites eocenica and S. montana. Although the actual genus to which these species belong may be in doubt, their reference to the palm famUy cannot be questioned. In the Uving flora, pahns are typicaUy tropical and subtropical in distribution. Their most northerly representatives range only into the warm temperate lowlands of the Atlantic Coast {Sabal pabnetto and Serenoa serrulata), CaUfornia {Washingtonia filamentosa), southern Europe (Chamcerops humilis), and southern Asia (Chamcerops spp. and Rhapis spp.). The Pandanacese, represented by the fossU species Pandanites corsoni, the Myrta- cese (Myrica torreyi), the DiUeniacese (Rhamnus cleburni), and the Lauraceae {Laurophyllum meeki, Ficus tesselata, Rhamnus salicifolius) are hkewise typicaUy tropical and subtropical famiUes whose modern distribution in temperate forests is restricted to the warm lowlands bordering the subtropics. The combined evidence presented from the study of the leaf characters of the Medicine Bow species, the distribution of closely related genera of the flora as a whole, and the climatic requirements of modern famUies which are represented in the fossU flora is consistent throughout in indicating warm temperate to subtropical FOX HILiLS AND LOWER MEDICINE BOW Table 3 29 Fossil Species Modern Correlatives Modern Distribution Cool Temp. Warm Terap. Sub- trop. Trop. Canna cf. C? magnifolia Carpites walcotti Celastrus taurinensis Cinnaraomum affine Cinnaraoraum linifolium Cissites lobatus Comophyllum wardii Dombeyopsis obtusa Dombeyopsis trivialis Dryophyllum subfalcatum Equisetum sp. Ficus planicostata Ficus cockerelli Ficus crossii Ficus tesselata Ficus cowanensis Grewiopsis saportana Juglans leconteana Juglans newberryi Canna . Thalia. Nyssa . Aralia .... Vibumum . Cinnamoraum . Cinnamomum Cocculus (Cebatha) . Lindera Cissus Pterospermum . Cornus Rhamnidium . . Berchemia .... Premna Ficus Cocculus (Cebatha) . Cocculus (Cebatha) . Cissus Quercus. . Castanea . Equisetum . Ficus Cocculus (Cebatha) . Hyperbsena Alchornea Mallotus Pterospermum Cissampelos Cocculus (Cebatha) . , Magnoha . . Nectandra . Magnolia . Laurus . . . Grewiopsis . . . Callichlamys . Maclura Apocynum . . . Juglans . . Sapindus. 30 C O N T R I B U T I O N S TO P A L ^E O N T O L O G Y Table 3 — Continued Fossil Species Modern Correlatives Modern Distribution Cool Temp. Warm Temp. Sub- trop. Trop. Laurus socialis? Magnolia nervosa Magnolia pulchra Magnoliophyllum cordatum Myrica torreyi Pandanites corsoni Phyllites colubrinoides Phyllites craigensis Phyllites trillioides Phyllites sp. Pistacia eriensis Rhamnus cleburni Rhamnus saUcifolius Sabahtes montana Sabahtes eocenica Sequoia nordenskioldi Trochodendroides nebrascensis Typha sp. Viburnum marginatum Viburnum montanum Vitis stantoni X X X X X X X X X X X X X X X X X X X X X X Persea X X X X X X Magnoha Magnoha X X X X X Pandanus Zizyphus X X X X X X X X X X Ilhcium Trillium X X X X X X X X X X X X X X Rhus Doliocarpus X X X X X X Davilla X X Persea Nectandra Sabal X X X X X X X X Sabal Seauoia X X X X X X Tvpha X X X X Viburnum . . . Vitis X Totals 26 48 67 41 FOX HILLS AND LOWER MEDICINE BOW 31 conditions of growth. Equally in harmony is thc ncgative evidence of the absence of the usual cool tempcrate genera and thc scarcity of conifers. CORRELATION In attempting to cstabUsh age relationships between other rccordcd floras and that of the lower Medicine Bow forraation, thrce basic assumptions are made. These assumptions are: (1) that tlie Sphenodiscus zone of the Fox Hills formation is of essentially the same age whcrever encountered; (2) that the Triceratops dino- saur zone is Kkewise essentially contemporaneous wherever found; and (3) that the Lance, Laramie, and lower Medicine Bow formations are conformable on the Sphenodiscus-hearing Fox Hills formation, and, thercfore, only shghtly younger than Fox Hills age. Since none of these assumptions seems, in my opinion, doubtful or unsupported, the age of the lower Medicine Bow flora is not open to serious question. Its occurrcnce in beds lying within 1300 feet conformably above the Sphenodiscus zone of the Fox Hills formation and 2500+ fect below Triceratops-hearing beds clearly estabhshcs the upper and lowcr hmits of its time-range. Whether either or both of these zones are referred to the Upper Cretaceous or the Paleocene is immaterial to the immcdiate problem of correlation. In the following analysis, therefore, the lower Medicinc Bow flora is regarded as a standard flora of known stratigraphic position with which othcr floras may bc advantageously compared or contrasted. As indicated in table 4, the Medicine Bow plant species of the present report total 64, of which 14 are either new or represcnted by specimcns insuflficicnt for comparison with previously recorded species. The rcmaining 50 spccies have bcen describcd or reported in othcr floras, which are discussed in the order of their similarity to the Medicinc Bow flora. Laramie flora — It is at oncc apparent from table 4 that the Laramie flora ^ of the Denver basin, Colorado, shows the closest relationship to the lower Medicine Bow flora, with 27 species in common. The samc conclusion was reachcd by Knowlton on the basis of smaUer, undescribed plant coUections obtained from scattered locahties in the Medicine Bow formation." This close relationsliip indi- cated by the floras is amply corroborated by several hncs of evidence: (1) Both the Laramie and the Medicine Bow formations occupy the same stratigraphic posi- tion conformably above the Fox Hills formation, which in both sections carries the diagnostic Sphenodiscus; ^ (2) both formations he conformably or with only local disconformity below beds yielding remains of Triceratops; (3) in the Medicine Bow formation the plant remains were obtained from the basal 1300 feet, which is approximately the same thickness of beds from which the Laramie plants wcre obtained; (4) the fresh-water invertebratc fauna recordcd from the Medicine Bow formation is of cssentiaUy the same character as that of the Laramie formation.* In my opinion there is thus sufficient unanimity of stratigraphic, palseontologic, ' Knowlton, F. H., U. S. Geol. Survey Prof. Paper 130, 1922. 2 Knowlton, F. H., ibid., 100. s Dobbin, C. E., and Roeside, J. B., Jr., U. S. Geol. Survey Prof. Paper 158, 23, 1930. * Henderson, J., Geol. Soc. Am., Spec. Pap. No. 3, 32, 1935. Table 4 — Local and outside distrihution of Medicine Bo 1« species Species * m CO >. >> 3 >. o 3 fO CJ ffl g ■5 1 1 'a 2 o k-. OJ > s 9 X X X 8 % ■a c "3 3 *3 o a Q o 3 ,o "4? > ? X ? X X X X X X X X X X X 12 c 3 a c "? 1 C3 < " 1 ci _M 0 0 £ X X X X X X X X X 8 1 CQ m X X X X X X X ? X X X X X ■ ■ X X X X X X X X 20 .S 03 ? X X X X X X X X X X X X X X X X X X X X X X X X X X X 27 0 0 c: s X X X X X X X X X X X X ? ? X X X X X X X X X X X X X X 26 ■a 3 1 1 tc 0 1 3 s 1 3 S 03 « ? X X X X ? X X X X X X X 11 d C3 0 L > c Q 7 X X X X X X ? X X X X X X ? X X X X X X 18 c 0 «1 c 03 > ? 7 0 03 c § X 0 0 2 c .0 "3 D 0 2 < 1 _ X X X X X X X X X X X X X X X X X X X X X X X X X X X ? X X X X X X X X X X X X X X Rrarta*' Canna of C "^ maenifolia X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X ? X X X X X X X X X X " X X X X X X X X X ? X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Trofhodendroides nebrascensia X X X X X X X X X X X X X X X Totals— 64 18 36 27 33 15 20 ' Species with outside distribution, 50. FOX HILLS AND LOWER MEDICINE BOW 33 and pakcohotanic evidcncc to indicate essential contemporancity between the Laramic formation of tlie Denver basin and at least the lower third of the Medicine Bow formation. Lance flora — Table 4 shows that there are 26 species in common between the Lance flora and that of the lower Medicine Bow formation. This is almost as great a similarity as between thc Laramie and the Mcdicine Bow fiora. Moreover, this similarit}' must by simple mathematical dcduction lcad to the conclusion that the Lance and Laramie floras are more nearly contcmporaneous than has ever previously been admitted. Knowlton maintained that the Lance flora was dis- tinctly youngcr than that of the Laramic/ although he admittcd that the Lance flora had not at that time bccn adequately describcd and that thcre was considcr- able uncertainty attached to some of the localities from which Lance plants had been collected. In an attempt to remedy both of thcse situations I have begun a detailed field and laboratory study of thc Lance flora, which to date has yielded the following preliminary observations : (1) Of the 58 rccordcd localitics ^ of reputed Lancc plants, 14 localitics are definitely at horizons in thc Fort Union formation, 8 are definitely and 7 probably in the TuIIock or Ludlow membcr of the Lance, 11 are definitely and 9 probably in the Hell Creek member of the Lance, 1 is defi- nitely in the Colgate member of the Fox Hills sandstone, and the remaining 8 localities are of very doubtful stratigraphic or gcographic position; (2) the list of Lance plant species publishcd by Knowlton in 1919 ^ was at that date ncither complete nor reliable, and must be completely revised to be of scientific value; (3) the trcatment of the Lance flora as a single unit flora will lead to further con- fusion; pending a more thorough invcstigation, it is apparent even in a preliminary enumeration of species that the flora of the HcII Creek member of the Lance is more closely allicd to that of the Laramie, Medicine Bow, Colgate, etc, of the latest Cretaceous, whereas the floras of the TuIIock and Ludlow members of the Lance are of Fort Union (Paleoccne) facies; (4) the only thoroughly reliable infor- mation regarding the Lance flora is that of Berry * and Brown ^ on floras from the Hell Creck mcmbcr; ncithcr of these floras includes elemcnts of Fort Union aspect. Despite the confuscd state of the Lance flora, it has been possible to show a close relationship between the lower Medicine Bow flora and that of the Lance, mainly from the coUcctions of Brown and Berry at the U. S. National Museiun, and from my own collcctions at Princeton Univcrsity. AU of thcsc collcctions are from the lower or Hell Creek member of the Lance, suggesting esscntial contem- poraneity between the lower Lance, lower Medicine Bow, and Laramie formations. Here is convincing corroboration of our basic assumptions, since the lower Lance, Uke the Laramie and the lower Medicine Bow, is conformably just above the Fox ' Knowlton. F. H., op. cit.. 104. » Stanton, T. W., and Knowlton, F. H., Bull. Geol. Soc. Am., vol. 8, 127-156, 1897; Leonard, A. G., U. S. Geol. Survey BuU. 316, 201, 202, 1907; Knowlton, F. H., Washington Acad. Sci. Proc, vol. 11, 179-238, 1909; Knowlton, F. H., Jour. Geol., vol. 19, 369-371, 1911; Calvert, W, R,, U. S. Geol. Survey BuII. 471-D, 195, 198, 1912; Winchester, D. E., Hares, C. J., Lloyd, E. R., and Parks, E, M., idem, BuII, 627, 24, 25, 1916; Rogers, G. S., and Lee, W., idem, BuII. 749, 29, 34, 1923; Hewett, D. F., idem, Prof. Paper 145, 27-35, 1926; Hares, C, J., idem, BuII. 775, 29, 1928; Dobbin, C. E., idem, BuII. 812-A, 11, 1929; Berry, E. W., idem, Prof. Paper 185-F, 127-132. 1934; Pierce, W. G., idem, BuII. 847-B. 56, 1936. 'Knowlton, F. H., U. S. Geol. Survey BuU. 696, 764, 1919. * Berry, E. W., U. S. Geol. Survey Prof. Paper 185-F, 1934. 'Brown, R. 'W'., U, S. Geol. Survey Prof. Paper (Manuscript, 1937). 34 CONTRIBUTIONS TO PALiBONTOLOGY Hills sandstone yielding Sphenodiscus lenticularis,^ and is below or within the zone of abundant Triceratops remains. Further discussion of the stratigraphic and floral problems of the Lance will be presented in later papers. Black Buttes flora — With 20 species common to the Medicine Bow formation, the flora from near Black Buttes in southwestern Wyoming is apparently com- parable, though not necessarily exactly contemporaneous. The Black Buttes flora occurs in the so-called "Laramie" formation (Black Buttes coal group),- which in this region is said to lie stratigraphically just above the Lewis shale. Knowlton has given an excellent historical review of the work on this flora.^ Although he at one time considered the flora of true Laramie age and equivalent to that of the Hell Creek member ("Ceratops beds") of the Lance,^ Knowlton in his later years was incHned to regard it as younger than the Laramie, and considered the beds to Ue unconformably on the massive sandstones (Fox Hills?) at the top of the Lewis shale.^ I visited the Black Buttes region during the summer of 1936 and failed to find any evidence of the unconformity which Knowlton had proposed. Until a complete restudy and revision of the stratigraphy, palseontology, and palseobotany of this section is completed, it will be impossible to make a definite age determina- tion of the flora. The high degree of similarity to the floras of the lower Medicine Bow formation cannot, however, be overlooked, and suggests similar age for at least some of the plant-bearing horizons of the Black Buttes group. Denver-Dawson floras — There are 17 species in common between the floras of the Denver and associated formations ^ and those of the lower Medicine Bow. This is considered a sufficiently large number of species to indicate that at least a part of the Denver-Dawson beds are of lower Lance-Laramie age. Until a few years ago such a statement might have been considered absurd, since it was supposed that both the Dawson and the Denver (with its basal Arapahoe) were separated from the underljdng Laramie by a great unconformity representing a hiatus of sufficient length to remove 14,000 feet of sedimentary rocks from the mountains to the west of the region.'' Dane and Pierce ^ have recently shown that the Arapahoe-Denver sequence is equivalent in position and presimiably in age to the Dawson, and that the great unconformity is at best of local stream-channel origin. It is, therefore, not unreasonable to expect Laramie species to extend into the Denver or Dawson. Of the 17 lower Medicine Bow species which occur in the Denver-Dawson, only four species in fact do not occur also in the underlying Laramie. There are still many apparent differences between the large Denver-Dawson floras and those of the lower Lance, lower Medicine Bow, and Laramie formations. As Dane and Pierce have indicated, "the upper parts of the Dawson and Denver may possibly include beds of Eocene age." ^ More careful coUecting will be neces- ' Dobbin, C. E., and Reeside, J. B., Jr., U. S. Geol. Survey Prof. Paper 158, 20, 1930. 2 Schultz, A. R., U. S. Geol. Survey BuU. 702, 22-24, 1920. ' Knowlton, F. H., U. S. Geol. Survey Prof. Paper 130, 62, 1922. ' Knowlton, F. H., BuU. Geol. Soc. Am., vol. 8, 136, 143, 156, 1897. ' Knowlton, F. H., U. S. Geol. Survey Prof. Paper 130, 62-65, 1922. « Knowlton, F. H., U. S. Geol. Survey Prof. Paper 155, 1930. ' Emmons, S. F., Cross, W., and Eldridge, G. H., U. S. Geol. Survey Mon. 27, 209, 1896. » Dane, C. H., and Pierce, W. G., BuU. Am. Assoc. Petr. Geol., vol. 20, no. 10, 1936. » Dane, C. H., and Pierce, W. G., ibid., 1328. FOX HILLS AND LOWER MEDICINE BOW 35 sary in this region bcfore the details of stratigraphic position, associated vertebrates, invertebratcs, and plants can be uscd to determine the true sequence and age relationships. Vermejo-Raton floras — The Vermejo and Raton floras were described by Knowlton ^ as two distinct floras occurring in beds which were separated by a definite unconformity. Although I have not visitcd this region, it secms apparent from the pubUshed descriptions of sections that the evidence for the unconformity can in most cases be interpreted as due to local stream-channeling. Table 4 shows that there are 12 and 11 Medicine Bow species in common with the Vermejo and Raton floras, respectively. In my opinion, this indicates that the unconformity, if it does exist, represents very little if any hiatus. If the Vermejo-Raton floras are combined they possess 20 species in common with the flora of the lower Medicine Bow formation, suggesting that at least a portion of the Vermejo-Raton sequence is of lower Medicine Bow age. In the Vermejo flora there are, to be sure, floral elements that suggest an age slightly older than Laramie or lower Medicine Bow. This is consistent with the upper Pierre rather than Fox Hills age assignment of the Trinidad sandstone, which lies conformably below the Vermejo. On the other hand, there are a number of Paleocene species in the Raton floras, suggesting that, as in the case of the Denver-Dawson beds, the lower portion of the Raton formation may be Cretaceous and the upper portion Paleocene and possibly younger. Fort JJnion flora — The extensive Fort Union flora of the Rocky Mountain region is well known though in need of critical restudy, which is at present being undertaken by Dr. Roland W. Brown of the U. S. Geological Survey. Knowlton ^ regarded the Lance and Fort Union floras as essentially similar, a view which has often been quoted in geological reports and textbooks. In view of the close rela- tionship now estabUshed between the lower Lance and the lower Medicine Bow floras, it is therefore surprising to find only two species in common between the lower ]\Iedicine Bow and Fort Union floras. This somewhat anomalous situation is in part explained by the discovery, as discussed under the Lance flora above, that some of KnowIton's so-called Lance collections were obtained from strata now conclusively known to be within the Fort Union formation. It has also become apparent that the lower Lance (Hell Creek) flora is quite distinct from that of the upper Lance (TuIIock and Ludlow) . The former is clearly of Laramie-Iower Medi- cine Bow aspect, whereas the latter is more closely related to the flora of the Fort Union. If this tentative observation is substantiated by further studies of the Lance floras which I have already under way, it brings into harmony the evidence of the plants and the vertebrates in placing the greatest faunal and floral break at the end of lower Lance (Hell Creek) time.^ This is still not in agreement ■nath the evidence afforded by marine invertebrates, since the Cannonball member of the Lance, which overlies the Hell Creek member, has yielded a fauna which is said > Knowlton, F. H., U. S. Geol. Survey Prof. Paper 101, 223-349. 1917. «Knowlton. F. H., Washington Acad. Sci. Proc, vol. 11, 218-226, 1909; Bull. Geol. Soc. Am., vol. 25, 334, 1914; U. S. Geol. Surv. Prof. Paper 155, 10, 1930. ' Since the completion of this manuscript, collections have been made in the lower Lance formation {"Ceratops beds") at its type locality. The fiora of about 75 species is closely related to the Medicine Bow and Laramie fioras and has little in common with the TuIIock, Ludlow, and Fort Union floras. 36 C O N T R I B U T I O N S TO P A L iE O N T O L O G Y to show a closer relationship to the underlying Fox Hills fauna of the Upper Creta- ceous than to any known early Tertiary fauna.^ Summary — The evidence of the fioras here discussed substantiates to a con- siderable degree the correlation of late Cretaceous and early Tertiary formations suggested previously by other Hnes of evidence. Figure 8 shows the generaUy N.W. Colo. S.W. Wyo. S. Wyo. S. Colo. E. Colo. E. Wyo., E. Mont, and Dakotas O s WASATCH RIVER \ WASATCM i FERRIS V POISON CANYON RATON \ 7 >STL£ ROCK WASATCH .POST-,, LARAMIE 2 / 1 WASATCH 21J za: pu / y MEDICINE BOW rONCUE R^VER Z o 2 D h a o b- uPOST-,, LARAMIE Z o o q: u > z u ^ LARAMIE v' -fir---- -^ LEBO □ ARA- PAHOE CANNON^ 6ALL, TULLOCK AND UJDLOW UJ O z < J S o 1 g OJ s Oh 3 MEDICINE BOW/ LARAMIE HELL CREEK FOX HILLS FOX HILLS -.,,_ VERMEJO _,-'"' FOX HILLS ICOLGATE FOX HILLS PIERRE PIERRE LEWIS / LEWIS \ LEWIS -- -^ TRINIDAD ^ PIERRE ALMOND WILLIAMS FORK ERICSON MESAVERDE \ _J FiQ. 8 — Correlation chart showing proposed changes by dotted lines and arrows. accepted correlation chart of the formations involved, modified from Miss Wil- marth's charts as suggested by the U. S. Geological Survey.^ Further proposed changes based on the present floral studies, indicated by dotted lines and arrows, are the Laramie age of part of the Vermejo-Raton sequence and the Fort Union age of the Cannonball, Tullock and Ludlow members of the Lance formation. The present study supports the conclusion of Dane and Pierce that the lower Dawson and the Arapahoe-Iower Denver should be referred to the Upper Cretaceous. SUMMARY The primary purpose of the present report is the description and discussion of the flora coUected from the terrestrial sedimcnts conformably overlying the 1 Stanton, T. W., U. S. Geol. Surv. Prof. Paper 128-A, 13, 15, 1920; Dane, C. H., and Pierce, W. G., Bull. Am. Assoc. Petr. Geol., vol. 20, no. 10, 1327. 1936. 2 Dane, C. H., and Pierce, W. G., op. cit., 1327. FOX HILLS AND LOWER MEDICINE BOW 37 youngest mariiie zone of the Upper Cretaceous system of southern Wyoming and northwestern Colorado. In southern Wyoming the plant remains were coUected from 4 localities in the Hanna and Carbon basins. Here the plant horizons Hc within the lower third of the Medicine Bow formation. Measured sections indicate the exact positions of the plant-bearing beds with reference to the marine Sphenodiscus zone of Fox Hills age. Evidence is prcsented for the recognition of thc Fox Hills formation, on both lithologic and palseontologic grounds, in the strata formerly referred to the upper Lewis or basal Mcdicine Bow formations. In northwestern Colorado the plant remains were obtained from 2 locahties near Craig from bcds prcviously referrcd to the "Laramie" formation. The basal sandy beds of this formation have yielded marine invertebrates, including the diagnostic Fox Hills ammonite, Sphenodiscus. On the basis of stratigraphic, Utho- logic, and palseontologic evidence the section is divided into the Medicine Bow formation above and the Fox HiUs formation below, the latter lying conformably above the Lewis shale as in southern Wyoming. The plant remains from the 6 localities are treated as a unit, referred to as the lower Medicine Bow flora. Of the 64 recognizable forms 47 are referred to species previously described elsewhere, 11 are described as new species, and 6 are recorded as not specifically determinable. It is recognized that many of the generic references are of doubtful botanic validity. Although more plausible generic alter- natives are suggested for most of the species, few changes in nomenclature have been made because of the lack of corroborative proof at the present time. In some species the large suites of specimens obtained have made possible new descriptions and re\asions of synonymous forms. From a quantitative viewpoint 11 species are shown to be dominants on the basis of abundance of individual leaves and occurrence at the largest number of localities. The structural characters of the dicotyledonous leaves in the flora indicate warm temperate to subtropical conditions of growth, more nearly approaching the latter. SimUar climatic impUcations are shown by the modern distribution of famiUes and genera whose systematic positions appear certain, and of generic alternatives suggested by simUarities between the fossU leaves and those of living genera. The age of the lower Medicine Bow formation is rather definitely estabUshed by its stratigraphic position, its relation to the Sphenodiscus zone and to Triceratops- bearing beds, and its associated fresh-water molluscs. The Laramie and lower Lance (HeU Creek member) floras are shown to be most nearly simUar to that of the lower Medicine Bow and are considered essentially contemporaneous. Tliis conclusion is supported by the relation of both the Laramie and the lower Lance formations to the Sphenodiscus zone and Triceratops-hearing beds. From a pre- liminary study of the Lance floras it seems evident that the Hell Creek flora is essentiaUy of Laramie-lowcr Medicine Bow aspect (Upper Cretaceous) and the upper Lance (TuIIock and Ludlow members) flora is of Fort Union (Paleocene) character. The latter reference is in harmony with the evidence of the vertebrate 38 CONTRIBUTIONS TO PAL^ONTOLOGY faunas but is contradictory to the reputed Upper Cretaceous age of the Cannonball marine member of the Lance, which corresponds to the Tullock member in its position above the Hell Creek. Other formations in the Rocky Mountain region whose floras are shown to be comparable in character and age to that of the lower Medicine Bow formation are the so-called "Laramie" of Black Buttes, Wyoming and part of the Denver- Dawson beds and of the Vermejo-Raton sequence of Colorado. The extensive Fort Union flora of the Rocky Mountain region is considered to have little in common with that of the lower Medicine Bow. The often-quoted statement that the Fort Union and Lance floras are essentially similar in aspect is shown to be not strictly true. SYSTEMATIC PAL^OBOTANY GENERAL PROCEDURE In view of the preponderance of leaves in nearly all occurrences of fossU plant remains, it is only natural that the governing principle for taxonomic work is that the leaves of trees and shrubs are sufficiently well-defined, diagnostic, and constant within genera and species to make identification possible. There are, of course, exceptional cases such as the modern Sassafras officinale Nees and Ebermaier, whose leaf form varies normally from entire to simply lobate to doubly lobate. Fossil specimens of such variable leaf forms of a single species might conceivably lead to their identification as three distinct species. Were it possible to obtain a large suite of such specimens, however, showing intergradations be- tween the leaf forms, they would properly be designated a single species. Modern palEeobotanic technique clearly recognizes leaf variabihty within certain limits of both modern and fossU species. In a thousand leaves from a single modern tree there may be found a dozen variations in size or shape or other leaf characters. Systematically arranged, however, these variants usually are constant in other characters, such as margin, venation, or character of the base or tip. In deahng with fossU specimens the normal variants of a single species can likewise be recog- nized, provided that a sufficient number of specimens are coUected and studied. To achieve the most rehable identifications, therefore, it is desirable to obtain as large a suite of specimens of each fossil leaf type as is practicable. In the taxonomic study of the Medicine Bow floras the coUections were first sorted in the field into provisional species, based on similarity of leaf forms. From each suite of provisional species a representative suite of variants was next selected, if available, and shipped for laboratory study. Each selected specimen was then compared and contrasted with pubhshed descriptions and illustrations of the numerous Upper Cretaceous and early Tertiary species of North America. Where simUarities scemed unquestionable the specimens were compared with the type specimens, nearly all of which are in the coUections of the U. S. National Museum. In the larger suites of provisional species, it lias often been possible to sliow that certain leaf types previously regarded as distinct species were merely variants of another species. In the case of Ficus planicostata, for example, I was faced with F O X n I L L S A N D L O W K K M K D I C I N p: B O W 39 the alternativc of rcferrinG; the hirj^e suite of spccimens of this leaf type either to 8 spccies or to 1 spccies. The lattcr choice was inevitable, owing to the intergrada- tions bctween the numerous variants. Such a course naturally has lcd to a large synonjTny for this species as wcll as others whose leaves were present in large numbers. As a consequence, however, it follows that the spccics representcd by a large nunibcr of specimens are usually more rcliably identificd, and long synon^-mies of such species tend to add to their value for both stratigraphic and purely botanical purposes. In cases where only a few specimens of a given leaf type were obtained in the collections, the specific identifications are somcwhat less positive. In some cases, for example, the lack of preservation of some integral portion of a leaf has pre- vented comparison with some previously described species. Final disposition of such species must of necessity await the possibiUty of obtaining larger suites of more complete specimcns. The generic rcferenccs of many of the Medicine Bow species are open to serious doubt. Unless whoUy rehablc evidence to the contrary has been obtained, no new generic names have been substituted for those previously applied to the species. I-n many cases herbarium studies have convinced me that the generic references are unquestionably incorrect, yet my study of thousands of herbarium sheets of othcr genera have failed to reveal comparable leaf types. Although older generic refer- ences are thus retained for convenience and for lack of positive evidence for changes at the present time, it should be understood that such references have little botani- cal significance and few if any physical conclusions can be drawn from them. It is, of course, possible that rnany of the Upper Cretaceous genera have failed to survive in modern forests. It seems more likely, however, in view of the close similarity of some of the leaves and seeds to modern forms, that a continued search among modern plants will ultimately bring out the true botanic relationships. The few remains of seeds obtained in the collections have thus far been stubbornly unidenti- fiable, or at best determined only within a limit of 6 to a dozen generic possibilities. Possibly future determinations of petrified wood and spores may open the way to unsuspected relationships. In most of the generic references previously made by other writers, their temperate zone experience has been clearly manifest. The Medicine Bow flora contains a number of distinctly warm temperate and subtropical elements, and most of the genera suggested as alternatives to present references are of low latitude families. It is, therefore, becoming increasingly evident that a better acquaintance with more southerly forests will be necessary to a proper taxonomic study of the fossil species. During the winter of 1931, I spent 3 months in the West Indies and South and Central America studying the modern forests with Dr. Ralph W. Chaney of the University of California, under the auspices of the Carnegie Institu- tion of Washington. I have planned to extend these studies northward into Mexico and the Gulf states during the coming year. It is hoped that these experiences will lead materially to a bctter understanding of both the taxonomy and the ecologj' of the late Cretaceous floras of the Rockj' Mountain region. 40 CONTRIBUTIONS TO PAL^ONTOLOGY In the discussion of most of the species in the Medicine Bow floras, suggestions have been made, where possible, regarding what appear to be the more Hkely family and generic relationships. In several new species, however, pending more conclusive evidence, the leaves have been referred to the form genus Phyllites and the seeds to Carpites. Systematic descriptions of previously described species have been amended and new illustrations added where better material has been obtained. SYNONYMS AND CHANGES OF NAMES In the foUowing Ust all names which have been changed or synon^Tnized previ- ously by other writers are inchided with those of the present report. The justifica- tion for changes in species concepts is based on either more adequate or more niunerous specimens, or on closer comparison with modern leaves. A fuU discussion and complete bibUographic references of synonymous foi-ms are given under each species in the systematic descriptions. Aleurites eocenica Lesquereux = Greiviopsis saportana Lesquereux. Alnus auraria Knowlton and Cockerell = Quercvs viburnifolia Lesquereux. Aralia taurinensis (Ward) Sanborn = Celastrus taurinensis Ward. Asplenium magmim Knowlton of Hollick = Asplenium? coloradense Knowlton. Asplenium martini Knowlton = Aspleniumf coloradense Knowlton. Berrya racemosa Knowiton = Leguminosites arachioides jninor Berry. Betula fallax Lesquereux (in part) = Quercus viburnifolia Lesquereux. Castalia stantoni Knowlton = Vitis stantoni (Knowlton) Brown. Celastriis curvinervis Ward = Celastrus taurinensis Ward. Celastrus gaudini Lesquereux (in part) = Quercus viburnifolia Lesquereux. Celastrus ovatus Ward = Celastrus taurinensis Ward. Celastrus wardii Knowlton and Cockerell = Celastrus taurinensis Ward. Cinnamomum affine Lesquereux (in part) = Ficus trinervis Knowlton. Cinnamomum sp. Knowlton = Ficus trinervis Knowlton. Cissusf cannoni Knowlton = Trochodendroides nebrascensis (Newberry) Dorf, n. conib. Cissus lesquereux Knowlton = Trochodendroides nebrascensis (Newberry) Dorf, n. comb. Cornus emmonsii Ward (in part) = Ficus crossii Ward. Cornus impressa Lesquereux of Knowlton = Ficus crossii Ward. Cornus prceimpressa Knowlton = Ficus crossii Ward. Cornus studeri Heer? of Lesquereux, Ward, and Knowlton = Cornophyllum wardii Dorf, n. sp. Daphnogene elegans Watelet of Ward = Laurophylhmi meeki Dorf, n. sp. Diospyrosf flcoidea Lesquereux (in part) = Laurophtjllum meeki Dorf , n. sp. Dryophyllum aquamarum Ward = Dryophyllum sxibfalcatum Lesquereux. Dryophyllum bruneri Ward = Dryophyllum subfalcatum Lesquereux. Dryophyllum falcatum Ward = Dryophyllum subfalcatuni Lesquereux. Dryophylhmi {Quercus) subfalcatum Lesquereux = Dryophyllum subfalcatum Lesquereux. Ficus clintoni Lesquereux = Ficus planicostata Lesquereux. Ficus haddeni Knowlton = Ficus trinervis Knowlton. Ficus impressa Knowlton = Ficus planicostata Lesquereux. Ficus irregularis (Lesquereux) Lesquereux (in part) = Ficus coloradensis Cockerell. Ficus latifolia (Lesquereux) Knowlton = Ficus cockerelli Knowlton. Ficus le^i Knowlton (in part) = Ficus planicostata Lesquereux. Ficus neoplanicostata Ivnowlton (in part) = Ficus planicostata Lesquereux. Ficus neoplanicostata Knowlton (in part) = Ficus trinervis Knowlton. Ficus planicostata clintoni (Lesquereux) Knowlton = Ficus planicostata Lesquereux. Ficus planicostata goldiana Lesquereux = Ficus planicostata Lesquereux. Ficus planicostala latifolia Lesquereux = Ficus cockerelli Knowlton. Ficus planicostata magnifolia Knowlton = Ficus planicostaia Lesquereux. Ficus prcelatifolia Knowlton = Ficus planicostata Lesquereux. FOX HILLS AND LOWER MEDICINE BOW 41 Ficus 'practrinervis Knowlton = Ficus trinervis Knowlton. Ficus pseudopopulus Lesquereux (in part) = Ficus planicostata Lesquereux. Ficus? smithsoniana (Lesquereux) Lesquereux = Rhamnus salicifolius Lesquereux. Flabcllaria communis Lcsquereux = Sabalites eocenica (Lesquereux) Dorf, n. comb. Flabellaria eocenica Lesquereux (in part) = Sabalites eocenica (Lesquereux) Dorf, n. comb. Fraxinus? princctoniana Knowlton = Dryophyllum subfalcatum Lesquereux. Fraxinus sp. Knowlton = Quercus viburnifolia Lesquereux. Grewiopsis eocenica (Lesquereux) Knowlton = Grewiopsis saportana Lesquereux. Grewiopsis ficifolia Ward = Grewiopsis saportana Lesquereux. Juglans laramiensis Knowlton = Rhamnus salicifolius Lesquereux. Leguminosites arachioides Lesquereux of Penhallow = Leguminosites arachioides minor Berry. Malapoenna louisvillensis Knowlton = Ficus trinervis Knowlton. Myrica coriacea Knowlton = Myrica torreyi Lesquereux. Myrica dubia Kjiowlton = Myrica torreyi Lesquereux. Myrica oblongifolia Knowlton = Myrica torreyi Lesquereux. Myrica sp. Knowlton = Myrica torreyi Lesqucreux. Myrica torreyi minor Lesquereux = Myrica torreyi Lesquereux. Nyssaf racemosa Knowlton = Leguminosites arachioides minor Berry. Palwoaster? similis Knowlton = Palceoaster inquirenda Knowlton. Phyllites populoides Knowlton = Donibeyopsis obtusa Lesquereux. Phyllites trinervis Knowlton = Dombeyopsis obtusa Lcsquereux. Pistacia hollicki Knowlton = Pistacia eriensis Knowlton. Platanus aceroides latifolia Knowlton (in part) = Viburnum marginatum Lesquereux. Platanus guillclmce heerii Knowlton = Viburnum marginatum Lesquereux. Platanus heeri Lesquereux of Ward = ^^iburnum marginatum Lesquereux. Platanus marginata (Lesquereux) Heer of Knowlton = Viburnum marginatum Lcsquereux. Platanus platanoides (Lesquereux) Knowlton (in part) = Vibiirnum marginatum Lesquereux. Populus arctica Heer of Lesquereux (in part) = Trochodendroides nebrascensis (Newberry) Dorf, n. comb. Populus crenata Ungcr of Lesquereux (in part) = Trochodendroides nebrascensis (Newberry) Dorf, n. comb. Populus denverensis Knowlton = Quercus viburnifolia Lesquereux. Populus jacksoni Knowlton = Trochodendroides nebrascensis (Ncwberry) Dorf, n. comb. Populus nebrascensis Newberry = Trochodendroides nebrascensis (Newberry) Dorf, n. comb. Populus nebrascensis var. grandidentata Lesquereux = Trochodendroides nebrascensis (Newberry) Dorf, n. comb. Popidus nebrascensis var. acute-dentata Lesquereux = Trochodendroides nebrascensis (Newberry) Dorf, n. comb. Populus nebrascensis var. longifolia Lesquereux = Trochodendroides nebrascensis (Newberry) Dorf , n. comb. Populus nebrascensis var. rotundata Lesquereux = Trochodendroides nebrascensis (Newberry) Dorf , n. comb. Populus? neomexicana Knowlton = Dombeyopsis obtusa Lesquereux. Populus tenuinervata Lesquereux = Trochodendroides nebrascensis (Ncwberry) Dorf, n. comb. Populus zaddachi Heer of Lesquereux (in part) = Trochodendroides nebrascensis (Newberry) Dorf , n. comb. Populus zeilleri (Lesqucreux) Knowlton = Trochodendroides nebrascensis (Newberry) Dorf, n. comb. Protoficus zeilleri Lesquereux = Trochodendroides nebrascensis (Newberry) Dorf, n. comb. Pterospermites neomexicanus Knowlton = Ficus planicostata Lesquereux. Pterospermites nervosus Knowlton = Magnolia nervosa (Knowlton) Brown. Quercus doljensis Pilar of Ward = Dryophyllum subfalcatum Lesquereux. Quercus gracilis Newberry = Dryophyllum subfalcatwn Lesquereux. Quercus purdonensis Knowlton = Quercus viburnifolia Lesquereux. Quercus turbidenta Hollick = Dryophyllum subfalcatum Lesquereux. Quercus whitei Lesquereux of Knowlton = Quercus viburnifolia Lesquereux. 42 CONTRIBUTIONS TO PAL^ONTOLOGY Quercus whitmani Knowlton = Dryophyllum subfalcatum Lesquereux. Rhamnites knowltoni Berry (in part) = Ficus crossii Ward. Rhamnus belmontensis Knowlton and Cockerell = Rhamnus salicifolius Lesquereux. Rhamnus brittoni Knowlton = Rhamnus cleburni Lesquereux. Rhamnus elegans Newberry = Rhamnus salicifolius Lesquereux. Rhamnus rectinervis Heer of Lesquereux (in part) = Rhamnus cleburni Lesquereux. Rhus pseudo-meriani Lesquereux = Dryophyllum subfalcatum Lesquereux. Sabal communis Lesquereux = Sabalites eocenica (Lesquereux) Dorf, n. conib. Sabalf eocenica linowlton = Sabalites eocenica (Lesquereux) Dorf, n. comb. Sabal inquirenda Knowlton = Sabalites eocenica (Lesquereux) Dorf, n. comb. Sabal montana Knowlton = Sabalites montana (Knowlton) Dorf, n. comb. Sabal? rugosa linowlton = Sabalites eocenica (Lesquereux) Dorf , n. comb. Sabalites fructifer Lesquereux (in part) = Leguminosites arachioides minor Berry. Sabalites grayanus (Lesquereux) Lesquereux (in part) = Sabalites montana (Knowlton) Dorf, n. comb. Ulmus? irregularis Lesquereux = Ficus coloradensis Cockerell. Viburnum contortum Lesquereux? of Knowlton = Q^iercus vibiirnifolia Lesquereux. Viburnum? problematicum Knowlton = Viburnum montanum Knowlton. Vitis dakotana Berry = Vitis stantoni (Knowlton) Brown. Zingiberites? undulatus Lesquereux = Zingiberites dubius Lesquereux. Zizyphus coloradense Knowlton = Zizyphus hendersoni Knowlton. DESCRIPTIONS Family POLYPODIACE^ Genus ASPLENIUM Linn6 Asplenium? coloradense Knowlton (Plate 1, Figs. 1, 2) Asplenium? coloradejise Knowlton, U. S. Geol. Surv. Prof. Paper 101, 24.5, pl. 30, figs. 1, 2, 1917. Asplenium magnum Knowlton. IloIIick, Torreya, vol. 2, 146, pl. 4, figs. 1, 2, 1902. Asplenium martini ICnowIton, U. S. Geol. Surv. Prof. Paper 130, 111, pl. 2, fig. 6, 1922. Two well-preserved specimens and one fragmcnt are clearlyidentical with the figured and type specimens of Asplenium? coloradense Knowlton from the Vermejo formation of Colorado. Thc shape and disposition of the pinnules are precisely as in Knowlton's figure 2, and the vcnation is the same, consisting of about eight or nine pairs of veins which fork once just above the midvcin. Knowlton has previously included in this species the specimens from Rockvale, Colorado which Hollick had regarded as Asplenium magnum Knowlton. This seems clearly justifiable. It is not possible, moreover, to distinguish this species from Asplenium. martini Knowlton from the Laramie of Colorado. Although pointing out its close relationship, Knowlton maintaincd this as a separate species on the basis of "its larger size, more obtuse pinnules, and much stronger vena- tion." 1 In view of the variation shown in the shapes of the pinnules of Aspleniumf coloradense Knowlton, with those near the base being larger and more obtuse than the rest, and the identity of tlie venation, it seems impossible not to inchide all of these specimens in the single species which has priority. The generic reference is still open to question, pending the discovery of at- tached fructifications. Occurrence — Corson Ranch, Wyoming, Loc. P. 371. Colledion—V. C. Mus. Pal., Plesiotypes Nos. 1289, 1394. Genus DRYOPTERIS Andanson Dryopteris? carbonensis Knowlton (Plate 1, Figs. 3-7) Dryopterisf carhmensis I&owlton, U. S. Geol. Surv. Prof. Paper 130, 110, pl. 20, figs. 3-5, 1922. This form is represented by seven specimens, five of which are figured. Most of the speci- mens are fragmentary or poorly preserved. The general form, size, and disposition of tlie pinnules, however, is adequately shown in figure 6; details of tlie venation appear in figure 7. In all observable features these specimens are indistinguishable from the figures and type specimens of this Laramie species. The details of the venation are somewhat better preserved than in Knowlton's specimens, indicating, as he intimated, that some of the finer veins are once- forked near the midvein. Knowlton has previously recorded ^ the presence of this species in the Medicine Bow forma- tion in a manuscript Ust based on coUections at the U. S. National Museum. Unclassified speci- mens (Nos. 590, a, h, c, e) of similar pinnules have been identified in the coUections from Black Buttes, Wyoming, which I have consulted at the U. S. National Museum. There is no positive evidence for the reference of this species to the genus Dryopteris, except a general resemblance in shape, size, and venation of pinnules. A more definite allocation cannot be made without better material and associated fertile fronds. Occurrence — Corson Ranch, Wyoming, Loc. P. 371. Collection—V. C. Mus. Pal., Plesiotypes Nos. 1290, 1291, 1395, 1396, 1397. 'Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 130, 111, 1922. «Knowlton, F. H., U. S. Geol. Surv. Bull. 696, 752, 1919. 43 44 C O N TR I B UT I O N S TO PAL^ONTOLOGY Family SCHIZiEACEyE Genus ANEIMIA Swartz Aneimia sp. (Plate 1, Fig. 8) The single unsatisfactory specimen figured is the only one of this form in the coUections. It is clearly quite different from the other ferns but is unfortunately lacking in the characters essential for reliable identification. The size, shape, and arrangement of the pinnules seem to relate this form to several recorded fossil species of Aneimia (usually misspelled Anemia). It differs from A. elongata (Newberry) Knowlton, from the Laramie and Mesaverde formations of Colorado and Wyoming,' only in possessing crenate rather than coarsely toothed margins. It is equally close to A. mosbeyensis Ivnowlton from the Dawson formation ^ and A. siipercretacea HoIIick from the Vermejo.' The marginal crenations are also apparently the main difference between our specimen and the figure of A. occidentalis Ivnowlton from the Raton formation.'* I believe that if large, complete speci- mens of any of these species could be collected they might conceivably show all of the variations on a single frond which at present are considered specific differences. The i'eference of all of the specimens cited to the genus Aneimia seems open to question. In so far as leaf morphology is concerned they might equally well be referred to Asplenium, Poly- ■podium, or Pteris. In the absence of diagnostic fruiting structures, however, it is impossible to attempt a more accurate determination. Occurrence — Corson Ranch, Wyoming, Loc. P. 371. Collection—V. C. Mus. Pal., No. 1292. FamDy EQUISETACE.E Genus EQUISETUM Linn6 Equisetum sp. (Plate 1, Figs. 9, 13) The coUections contain a half dozen specimens which consist in part of ribbed, jointed stems and in part of tubers connected with jointed stems. The preserved characters do not seem to me to be sufficiently diagnostic to attempt a specific determination. The reference to Equisetum, however, seems rather certain. There appear to be very few figured specimens of comparable jointed stems from the late Cretaceous or early Tertiary of the Rocky Mountain region. It may be that our specimens might be comparable to several forms from the Lance or Denver, which are, unfortunately, listed simply as Equisetum sp.,^ and not figured. The tubers, on the other hand, are very similar to those of Equisetum perlcevigatum Cockerell from the Laramie formation of Colorado.^ Close resemblance of such generahzed portions of plants as underground tubers does not, however, seem to justify their being considered conspecific, or their being used for correlation purposes. Occurrence — Corson Ranch, Wyoming, Locs. P. 371, P. 372. Collectio7i—\J . C. Mus. Pal., Nos. 1293, 1294. > Knowlton, F. H., U. S. Geoi. Surv. Prof. Paper 130, 112, pl. 2, fig. 2, 1922. » Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 155, 28, pl. 8, fig. 9, 1930. ' Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 101, 248, pl. 30, fig. 5, 1917. * Knowlton, ibid., 285, pl. 54, fig. 2. ' Knowlton, F. H., U. S. Geol. Surv. BuII. 696, 259, 1919. « Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 130, 113, pl. 1, figs. 8, 9, 1922. FOX HILLS AND LOWER MEDICINE BOW 45 Family TAXODIACE^ Genus SEQUOIA Endlicher Sequoia? acuminata Lesquereux (Plate 2, FiR. 1) Seqxtoia acuminala Lesquercux, U. S. Geol. and Geog. Surv. Terr. Bull., vol. 1, 384, 1876; Rept. U. S. Geol. Surv. Terr., vol. 7, 80, pl. 7, figs. 15-16a, 1878; VKnowlton, U. S. Geol. Surv. Prof. Paper 130, 114, pl. 2, figs. 7, 8, 1922. There are only two incoinplete speciraens of this type in the coUections. They are suffi- ciently well defined, however, to indicate that they are not of the spreading, two-ranked leaflets here referred to Sequoin nordenskioldi Heer. The reference to S.? acuminata Lesquereux is made on the basis of identity with the figured and type specimens of the Laramie forms which Knowlton qucstionably referred to this species, and with the original types from Black Buttes, Wyoming. On the basis of material which I have collected from the lower Lance formation near Glenrock, Wyoming, I am prepared to present evidence in a forthcoming pubhcation that this species is not chstinct from Sequoia longifolia Lesquereux, a conclusion which both Lesquereux and Knowlton recognized as a possibihty in the reports cited above. After surveying herbarium materials of modern conifers, I am convinced that the reference to Sequoia is invahd. My studics of the better material from the Lance formation have as yet not revealed the true botanical relationship of the species. Occurrence — Corson Ranch, Wyoming, Loc. P. 372;northof Walcott, Wyoming, Loc. P. 373. Collection—V . C. Mus. Pal., No. 1295. Sequoia nordenskioldi Heer (Plate 1, Fig. 10) Sequoia nordenskioldi Heer, Fl. Foss. Arctica, vol. 2, pt. 3, 36, pl. 2, fig. 13b, pl. 4, figs. la, Ib, 4-38, 1870; Newberry, U. S. Geol. Surv. Mon. 35, 20, pl. 26, fig. 4, 1898. There are about a dozen specimens which are indistinguishable from Heer's original type figures of this species. The status of this species in North America is very unsatisfactory. In the first place, it cannot be separated from the widespread Tertiary leaf species Sequoia langsdorfii (Brongniart) Heer, which in turn is essentially similar to the Uving S. sempervirens Endhcher of the Pacific Coast. In the second place, cones of Sequoia found in the same Cretaceous strata as fohage referred partly to S. nordenskioldi, and partly to S. langsdorfii, have recently been described as S. dakotensis Brown.i These cones are comparable to those of the hving S. sempervirens and the Tertiary S. langsdorfii but are sufficiently different in significant characters to indicate that they belong to a distinct species, though not to Heer's S. nordenskioldi. It seems hkely, therefore, that the Sequoia fohage which here and elsewhere is referred to S. nordenskioldi is in reahty the foUage of the species whose cones are now called S. dakotensis. Until such time as the cones and foUage may be found in cUrect connection, however, it seems best to retain the widely used name S. nordenskidldi for the late Cretaceous fohage of this type. This species has been widely reported, though rarely figured, from the Lance and Fort Union of the Rocky Mountain region. I doubt if it wiU ever have the stratigraphic significance which can now be attached to species based on cone materiah Occurrence — Craig, Colorado, Loc. P. 375. Collection—U. C. Mus. Pal., Plesiotype No. 1296. Family TYPHACEiE Genus TYPHA Linn6 Typha sp. (Plate 1, Fig. 12) There are about a dozen weU-preserved fragments which are clearly referable to the genus Typha. The best of these, which is figured, shows the characteristic strong and weak paraUel veins crossed and connected irregularly by short, transverse veinlets. > Brown, R. W., Washington Acad. Sci. Jour., vol. 25. no. 10, 447, 1935. 46 CONTRIBUTIONS TO PAL^ONTOLOGY So few specimens of Cretaceous or Tertiary representatives of Typha have been described or identified specifically that there is nothing to be gained in describing the Medicine Bow specimens as a new species. From the environmental point of view, however, the presence of this common type of aquatic monocotyledon is of significance. Occurrence — Corson Ranch, Wyoming, Loc. P. 372. Colledion—V. C. Mus. Pal., No. 1297. Family PANDANACE^ Genus PANDANITES Dorf, n. gen. This generic name is proposed for specimens with the characteristics of the following type species. The name is derived from the family Pandanacese, whose members posscss spirally arranged Unear leaves which are leathery, keeled, and mostly spinulose on the margins and keel.^ Pandanites corsoni Dorf, n. sp. (Plate 3, Fig. 4) Only the single specimen figured was encountered in the collections. I •was at first inclined to regard this as the impression of a portion of a palm leaf . After careful study of all the available living and herbarium specimens of palms at The New York Botanical Garden, I was convinced that none of these was comparable to the fossil. It soon became apparent that its characters were very similar to those of the PandanaceiE. Description — Leaf of unknown lengt-h, width 5 cm., texture evidently leathery; prominent central keel, and each half of leaf marked by a subsidiary central dorsal keel ; leaf surface marked by 40 to 46 longitudinal, parallel veins which are irregularly connected by transverse veinlets which niay be perpendicular or obhque to the veins; margins of leaf invested with strongly spinose teeth about 5 to 6 mm. apart. The usual description of the leaves of the Pandanaceie emphasizes that both the leaf edges and the central keel are spinulose. In Uving specimens of Pandanus, however, I have noticed that the keel is perfectly smooth in the basal half of the leaves, becoming spinulose only in the distal half . The absence of teeth, therefore, in the fossil specimen is not considered of taxonomic importance. It is doubtful, moreover, whether spines along the keel would be discernible in a flat fossil impression. The modern PandanacetB are represented in the tropics and subtropics of the Old World by three genera: Sararanga Hemsl. (Solomon Islands) ; Freycinetia Gaudich (East Indies to Australia and New Zealand) ; and Pandanus Linne (Asia and Af rica) . They are usually f ound in lowland marshes or along sea-coasts and are not known to occur under typically temperate conditions. There are no fossil specimens known to me which are comparable to Pandanites corsoni. There is a superficial resemblauce in Bactrites pandanifolius Berry from the Eocene Lisbon forma- tion of Mississippi.^ This, however, is typically a flat leaf, with a single prominent midrib and longer teeth. This species is named for Mr. Andrew Corson, who was instrumental in the original dis- covery of fossil leaves on his property. Occurrence — Corson Ranch, Wyoming, Loc. P. 372. Collection—V. C. Mus. Pal., Holotype No. 1298. Family CYPERACE^ Genus CYPERACITES Schimper Cyperacites sp. (Plate 1, Fig. 11) There are a number of specimens in the collections which are quite obviously monocotyle- donous leaves with venation characters unlike those referred to Typha. As shown in the figured specimen they are long, 8 to 12 mm. wide, and provided with numerous fine, parallel veins about ■ Hutchinson, J.. The Families of Flowcring Planta, vol. II, 159, 1934. 2 Berry, E. W., U. S. Geol. Surv. Prof. Paper 92, 52, pl. 7, figa. 1-6, 1924. FOX HILLS AND LOWEIt MEDICINE BOW 47 .7 mm. apart; there appears to be a single finer voin between each pair of heavier ones; cross vcins are eithcr totally absent or vcry obscure. Comparable fossil remains are usually referred to the form genus Cyperacites. There is very Uttle significance to remains of this type cxcept to record the presence of addi- tional monocotyledons in the plant association. Occurrence — Corson Ranch, Wyoming, Locs. P. 371, P. 372; north of Walcott, Wyoming, Loc. P. 373; Craig, Colorado, Loc. P. 375. Collection—U. C. Mus. PaL, No. 1299. Family PALM/E Genus SABALITES Saporta Sabalites montana (Knowlton) Dorf, n. comb. (Plate 3, Fig. 2) Sabal monlana Knowlton, U. S. Geol. Surv. Prof. Paper 101, 2.53, pl. 32, %. 3, 1917; idem, Prof. Paper 130, 119, pl. 3, fig. 4, 1922. Sabalites grayanus (Lesquereux) Lesquereux (in part), Rept. U. S. Geol. Surv. Terr., vol. 7, 112, pl. 12, fig. 1 (not pl. 12, fig. 2), 1878; Knowlton, U. S. Geol. Surv. Prof. Paper 155, 36, pl. 9, fig. 5, 1930. The identification of the fossil palm leaves in the collections has led to considerable difficulty. The large size of the leaves usually makes it impossible to procure perfect or even near-perfect specimens. After an extended study of all available hving and herbarium specimens of palms at The New York Botanical Garden, I have conchided, however, that better taxonomic use can be made of fossil specimens in which the petiole and tlie adjoining basal rays are preserved. There are a number of features which are consistently uniform in any given genus or species and are consistently different in others. These features inchide (1) the character of the petiole: whether unarmed or armed with thorny spines; whether flat, convex, or concave on its upper surface; whether or not it is prolonged into a rachis; (2) the character of the rachis: short or long; straight, convex, or concave sides; (3) the approximate number of rays adjoining the rachis or petiole. With all of these features preserved, however, it does not seem advisable to refer the fossil forms to modern genera unless at least one single leaf is found completely preserved. The ending -ites can advantageously be added to modern famiUes or genera to indicate a close resemblance. In Sahalites, as here used, a resemblance to the subfamily Sabaleae is impUed. The present species is represented by a large number of fragmentary specimens, of which the figured specimen is the only one showing the basal portion of the blade. This cannot be dis- tinguished from the type or figured specimen of what Knowlton caUed Sabal montana from the Vermejo formation. In this species Knowlton justifiably included the specimens from the Mon- tana group (Mesaverde formation) of Point of Rocks, Wyoming, which Lesquereux had desig- nated as Sabalites grayamis. I am including also the specimen from the Denver formation which Knowlton referred to Sabalites grayantis. This is different in many respects from the type of S. grayanus (= Sahal grayana) from the Eocene of Mississippi,' but cannot be separated from Sabalites montana. This species is also recorded from the Laramie formation, as indicated in the above synonymy. The specimen from the Fruitland formation which Knowlton questionably referred to this species ^ is definitely of an entirely different type. Occurrence — Corson Ranch, Wyoming, Loc. P. 372; Elk Mountain road, Wyoming, Loc. P. 374. Collection—U. C. Mus. Pah, Plesiotype No. 1300. ' Lesquereux, Leo, Am. Philos. Soc. Trans., vol. 13. 412, pl. 14, figs. 4-6, 1869. 2 Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 98, 335, pl. 83, iig. 2, 1916. 48 CONTRIBUTIONS TO PAL^ONTOLOGY Sabalites eocenica (Lesquereux) Dorf, n. comb. (Plate 2, Fig. 6; Plate 3, Fig. 3) Flahellaria eocenica Lesquereux, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1872), 391, 1873; idem (1873), 380, 1874; idem (1876), 502, 1878; Rept. U. S. Geol. Surv. Terr., vol. 7, 111, pl. 13, figs. 1, 2 (not 3), 1878. Sabal communis Lesquereux, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1874), 311, 1876. Flahellaria comnmnis Lesquereux, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1875), 385, 1876. Sabal inquirenda Ivnowlton, U. S. Geol. Surv. Prof. Paper 101, 288, pl. 56, 1917. Sahal? rugosa Knowlton, U. S. Geol. Surv. Prof. Paper 101, 288, pl. 58, 1917. Sabal? eocenica Ivnowlton, U. S. Geol. Surv. BuII. 696, 557, 1919. The two figured counterparts and a number of fragmentary specimens are identical in every respect with the specimens originally called Flabellaria eocenica Lesquereux, and are indistinguish- able from the type and figured specimens of the other species cited above. Ivnowlton justifiably inckided all these species except Sabal? rugosa in his Sabalf eocenica. In a later report,' however, he confused the status of this species by referring it, along with many others, to Sabalites grayanus Lesquereux, originally described from the Eocene of Mississippi.^ After an examination of the type specimens of all of these species at the U. S. National Museum, I am convinced that Knowlton's Sabal? eocenica can readily be distinguished by the foUowing characters: (1) rounded and compressed aspect of the rays near their point of attachment; (2) number of folds — 60 to 80; (3) shghtly concave upper surface of the petiole; (4) truncate to rounded distal end of the rachis on its upper face; (5) prolongation to 6 to 8 centimeters of the abruptly narrowed, pointed, distal end of the rachis on its lower face. AU the specimens which Knowlton synonymized with S.? eocenica exhibit this set of characteristics. The type specimen of Sabal? rugosa, moreover, clearly indicates that it is the upper face of the same species. In view of the fact that I&iowlton questioned the reference to Sabal, it seems advisable to change the generic name to the less definitive Sabalites. As S. eocenica is now conceived, it is characteristic of the uppermost Cretaceous strata of the Rocky Mountain region, occurring in the Black Buttes, the lower Raton, the lower Medicine Bow, and the lower Lance floras.' Occurrence — Corson Ranch, Wyoming, Locs. P. 371, P. 372; north of Walcott, Wyoming, Loc. P. 373. Collection—H. C. Mus. Pal., Plesiotypes Nos. 1301, 1302. Genus PALMOCARPON Lesquereux Palmocarpon? subcylindrieum Lesquereux (Plate 2, Figs. 2, 3) Palmocarpon subcylindricum Lesquereux, Rept. U. S. Geol. Surv. Terr., vol. 7, 121, pl. 11, fig. 12, 1878. There are three good specimens of this type in the coUections. Lesquereux's description is as f ollows : "Fruit oblong or subcyhndrical, truncate at one end, spHt at the other in two diverging or slightly recurved, pointed lobes, distantly and obscurely veined toward the base." The Medicine Bow specimens fit this description perfectly except that the veining is less obscure. Comparison with the type specimens at the U. S. National Mu.seum indicated a close similarity in all characters except size. The veins are definitely not as far apart as Lesquereux's figures would indicate. Knowlton has reported additional specimens of this species from the Denver formation and has questioned the reference to Palmocarpon.^ 1 have no definite suggestion as yet regarding the true botanical affinity of the species. Occurrence — Corson Ranch, Wyoming, Loc. P. 371. Collection—V. C. Mus. Pal., Plesiotypes Nos. 1303, 1304. ' Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 155, 36, pl. 9, fig. 5, 1930. ' Lesquereux, Leo, Am. Philos. Soc. Trans., vol. 13, 412, pl. 14, figs. 4-6, 1869. ' Knowlton, F. H., Washington Acad. Sci. Proc, vol. 11, 204, 207, 1Q09. * Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 155, 42, 1930. FOX HILLS AND LOWER MEDICINE BOW 49 Family CANNACE^ Genus CANNA Linn^ Canna cf. C? magnifolia I^owlton (Plate 2, Fig. 5) The singlc incomplete specimen figured is the only onc of this kind in the coUections. It is evidently closcly related to Canna? magnijolia and Canna? sp. from the Vermcjo formation.' The only apparent difference is the absence of the faint alternation of thick and thin veins shown in Kno\vlton's figures and in the type specimens. Berry has reported fragments comparable to the same species from the Lance formation of South Dakota.^ I have seen Berry's specimens at the U. S. National Museum and find them indistinguishable from the observable features of the Medicine Bovv specimen. Among hving forms this type of leaf has been found to bear a general resemblance to the leaves of Musa and Heliconia of the Musacece, Maranta and Thalia of the Marantacea;, and Canna of the Cannacese. The uniform thickness of the numerous veins in the fossil specimens resembles most closely the leaves of Canna and Thalia. Occurrcnce — Corson Ranch, Wyoming, Loc. P. 372. Collectio7i—V. C. Mus. Pal., No. 1305. Family ZINGIBERACE.E Genus ZINGIBERITES Heer Zingiberites dubius Lesquereux (Plate 2, Fig. 4) Zing-ihcrites duhius Lesquereux, Rept. U. S. Geol. Surv. Terr., voL 7, 95, pl. 16, fig. 1, 1878; Ivnowlton, U. S. Geol. Surv. Prof. Paper 155, 43, 1930. Zingiherilcs? undulatus Lesquereux, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1873), 396, 1874. (Homo- nym, Heer, 1856.) The fragment figured is the only identifiable specimen of this type in the collections. It is clearly marked by parallel veins, the heavier ones about 1.5 to 2 mm. apart, with the lighter ones intermediate and numbering 5 to 7 between each heavier pair. These venation characters are similar to those of the specimens of Zingiberites dubius from the Denver formation. Both Lesquereux and Knowlton rightfuUy regarded this species as of doubtful taxonomic status. It is unfortunate that more complete specimens have not been found to add to the knowledge of its leaf form and characteristics. Occurrence — Corson Ranch, Wyoming, Loc. P. 372. Collection—JJ . C. Mus. Pal., Plesiotype No. 1306. Family MYRICACE.E Genus MYRICA Linn6 MjTica torreyi Lesquereux (Plate 4, Figs. 1, 2, 3) Myrica torreyi Lesquereux, U. S. Geol. Surv. Terr., 6th Ann. Rept., 392, 1873; Rept. U. S. Geol. Surv. Terr., vol. 7, 129, pl. 16, figs. 3-10, 1878; Ward, U. S. Geol. Surv., 6th Ann. Rept., 551, pl. 40, fig. 4, 1886; idem, Bull. 37, 32, pl. 14, fig. 5, 1887; Knowlton, U. S. Geol. Surv. BuU. 163, 34, pl. 6, figs. 1-3, 1900; idem, Prof. Paper 98, 90, 336, pl. 17, fig. 7, pl. 86, fig. 1, 1916; idem, Prof. Paper 101, 256, pl. 37, figs. 2-4, 1917; idem, Prof. Paper 130, 123, 1922. Myrica torreyi 7ninor Lesquereux, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1873), 397, 1874. Myrica coriacea Knowlton, U. S. Geol. Surv. Prof. Paper 101, 256, pl. 37, fig. 5, 1917. Myrica dubia Knowlton, U. S. Geol. Surv. Prof. Paper 130, 123, pl. 5, fig. 3, 1922. Myrica oblongifolia Ivnowlton, U. S. Geol. Surv. Prof. Paper 130, 124, pl. 21, fig. 1, 1922. Myrica sp., Knowlton, U. S. Geol. Surv. Prof. Paper 155, 44, pl. 11, fig. 7, 1930. Specimens of this type were present as one of the dominants in 4 of the 6 Medicine Bow localities. More than 50 well-preserved specimens were collected for study. At first I was ' Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 101, 254, 255, pl. 36, fig. 3, 1917. » Berry, E. W., U. S. Geol. Surv. Prof. Paper 185-F. 129, 1934. 50 CONTRIBUTIONS TO PAL^ONTOLOGY inclined to separate these into two species, Myrica torreyi and M. oUongifolia, on the basis of dif- ferences in shape, the forraer with narrowly lanceolate leaves and the latter witla long-elliptic to ovate-lanceolate leaves. I am now convinced, however, from gradations in shape which are apparent in the Medicine Bow collections and in various Upper Cretaceous collections at the U. S. National Museum, that all of these specimens belong to a single species whose leaf characters are amazingly consistent even though somewhat variable in size and shape. In view of the large amount of new material collected, it seems desirable to re-describe the species here. Description {Swpplementary) — Leaves linear to narrowly lanceolate to long-elliptic, sym- metrical, narrowed gradually to a long-cuneate base and a narrowly acute tip; length variable from 5 to 15 cm., width from 1 to 5 cm.; average dimensions appear to be about 11 cm. by 2.5; greatest width at the middle of the leaf or slightly below; petiole relatively short, thick, averaging about 2 cm. in length; midrib prominent on the lower surface of the ieaf; secondaries usually inconspicuous, numerous, and closely spaced, leaving the midrib at obtuse angles varying from 50° to 80°, averaging about 70°; secondaries irregularly forked, unbroken, or irregularly joined toward the margin where they meet a prominent intramarginal vein which borders the entire leaf 1 to 2 mm. from the edge; indistinct veins join the intramarginal vein to the closely spaced marginal serrations, which may be obscured by sUght curling of the leaf margin; tertiary venation obscure; texture coriaceous. Myrica toireyi, as previously conceived by various authors, is widely distributed in the late Cretaeeous sediments of the Rocky Mountain region. It has previously been recorded, usually abundantly, in the Black Buttes, Lance, Laramie, Mesaverde, Fruitland, Fox Hills, and Vermejo floras. Its presence in the Denver and Dawson floras is indicated by the specimens called Myrica sp. and M. torreyi minor which are clearly variants of the normal leaf type. Likewise, Myrica coriacea from the Vermejo formation and M. dubia and M. oblongifolia from the Laramie, which were described as closely similar to M. torreyi, are obviously only variants, a view which seems the more likely because of the previous record of the normal M. torreyi from these same beds. This species is an important and diagnostic index species of the late Cretaceous. It has never, to my knowledge, been reported from beds of undisputed Paleocene or younger age. Its characteristics, moreover, are usually well defined and not likely to cause confusion with other types. The reference of this fossil species to the genus Myrica is open to question. I have consulted all available herbarium sheets of both eastern and western hemisphere species of Myrica at The New York Botanical Garden, and have failed to find any species whose leaves are comparable to the fossil specimens. The pronounced intramarginal vein observed in the fossil species is never present in the leaves of Myrica; neither are the fine teeth nor the closely set secondary veins. The leaves of several genera of the Myrtacese, on the other hand, are comparable in most respects, differing only, in some cases, in the absence of fine marginal serrations. In the genera Amomis, Colycolpus, Calyptranthes, Eugenia, and Jambosa, for example, most of the species possess leaves of the same general shape and almost identical venation as in the fossil specimens. The closest resemblance observed is with Eugenia jambolana Lam. (sheet No. 5601) and E. {Phyllncalyx) sp. (sheet No. 7559) of Brazil, and with Jambosa jambos (L.) Millsp. (sheet No. 15020) of Colombia; these differ from the fossil leaves only in the absence of marginal serrations. It seems Ukely that Myrica torreyi will ultimately be referred to MyrtophyUum Heer, indicating resemblance to the family Myrtaceae. Occurrence — Corson Ranch, Wyoming, Loc. P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374; Craig, Colorado, Loc. P. 375. Collection—V. C. Mus. Pal., Plesiotypes Nos. 1307, 1308, 1309. Family JUGLANDACEiE Genus JUGLANS Linn6 Juglans leconteana Lesquereux (Plate 3, Fig. 1; Plate 4, Fig. 5) Juglans leconteana Lesquereux, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1870), 382, 1872; idem (1876), 517, 1878; Rept. U. S. Geol. Surv. Terr., vol. 7, 285, pl. 54, figs. 10-13, 1878; Knowlton, U. S. Geol. Surv. Prof. Paper 130, 121, pl. 8, figs. 1-3, 1922. This species is represented by 5 complete specimens and several fragments. They are indistinguishable from the specimens figured by Lesquereux and Knowlton from the Laramie and FOX HILLS AND LOWER MEDICINE BOW 51 Black Buttcs floras. They are similarly indistinguishable froin the original types consulted at the U. S. National Museum. These were apparently misplaced at the time Knowlton wrote Professional Paper 130, but have since been located. Both Lesquereux and Knowlton have considered the possibility that these leaves might represent variants of Juglans rugosa Lesquereux or J. rhamnoides Lesqucrcux. The continued discovery of leaves of the J. leconteana type, and the absence of leaves of the formcr types in the same strata tend to corroborate the maintenance of the latter species as distinct. The reference to the genus Juglans seems to be based on insufRcient evidence. I have noted a close resemblance of the speciraens to the leaves of Maclura (Moracese), especially M. aurantiaca Nutt., and Apocijnum (Apocynaceffi), especialiy A. pubcscens (?). Until some positive grounds are estabhshed for a change in name it seems best, however, to retain the old one. Occurrence — North of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374. Collection—V. C. Mus. Pal., Plesiotypes Nos. 1310, 1393. Juglans newberryi Knowlton (Plate 4, Fig. 4) Juglans neirberryi Knowlton, U. S. Geol. Surv. Prof. Paper 130, 120, pl. 20, figs. 8-10, 1922. The single specimen figured is the only one of this type in the collections. It is identical with the specimens of this species from the Laramie formation. Although Knowlton described the margins of his specimens as entire, the type specimens at the U. S. National Museum are some- what obscure along the margins, like the specimen here discussed, and appear to have a few fine serrations in the upper half of the leaf. These leaves resemble somewhat those of Juglans crossii Knowlton from the Green River formation,' differing mainly in the marginal characters and the more closely spaced secondary veins. The refercnce to Juglans may be correct, although several species of Sapindus have leaves which are virtually identical. Occurrence — Elk Mountain road, Wyoming, Loc. P. 374. Collection—lS. C. Mus. Pal., Plesiotype No. 1311. Juglans prserugosa Knowlton (Plate 4, Fig. 6) Juglans prccrugosa Knowlton, U. S. Geol. Surv. Prof. Paper 130, 121, pl. 5, fig. 2, pl. 21, fig. 5, 1922. This type of leaf is represented only by the complete specimen figured and several fragments. I was at first inclined to refer these to Juglans denveriana Knowlton ^ or J. rugosa Lesquereux ^ but find what seem to be important differences in shape, margin, and venation. In these characters the specimen is indistinguishable from the type and figured specimens of J. prarugosa from the Laramie, and fits its description perfectly. Knowlton has already reported the occurrence of this species in the Medicine Bow formation of the same region from which my collections were made.^ I have not bcen able to match these specimens with any modern leaves. Occurrence — ?Corson Ranch, Wyoming, Loc. P. 371; Craig, Colorado, Loc. P. 376. Collection—\] . C. Mus. Pal, Plesiotype No. 1312. Family FAGACEiE Genus DRYOPHYLLUM Debey Dryophyllum subfalcatum Lesquereux (Plate 5, Figs. 1, 2, 6) Dryophyllum (Quercus) suhjalcalum Lesquereu-X, U. S. Geol. and Geog. Surv. Terr. BuU., vol. 1, 379, 1876; idem, Ann. Rept. (1874), 301, 1876; Rept. U. S. Geol. Surv. Terr., vol. 7, 163, pl. 63, fig. 10, 1878; Knowl- ton, U. S. Geol. Surv. Bull. 163, 41, 1900; Washington Acad. Sci. Proc, vol. 11, 211, 1909; Brown, U. S. Geol. Surv. Prof. Paper, manuscript, 1937. Dryophyllum aquamarum Ward, U. S. Geol. Surv., 6th Ann. Rept., 551, pl. 37, figs. 3-5, 1886; idem, Bull. 37, 26, pl. 10, figs. 2-A, 1887; Knowlton, U. S. Geol. Surv. Prof. Paper 101, 299, pl. 70, fig. 2, 1917. > Lesquereux, Leo, Rept. U. S. Geol. Surv. Terr., vol. 7, 289, pl. 58, fig. 1 ( = J. denticulala Heer), 187S. 2 Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 155, 44, pl. 12, figa. 1, 2, pl. 13, figs. 2-t, 1930. s Lesquereux, Leo, Rept. U. S. Geol. Surv. Terr., vol. 7, 286, pl. 54, figs. 5, 14, pl. 55, figa. 1-9, 1878. * Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 130, 100, 1922. 52 CONTRIBUTIONS TO PAL^ONTOLOGY Dryophyllum bruneri Ward, U. S. Geol. Surv., 6th Ann. Rept., 551, pl. 36, figs. 6-9, 1886; idem, BuU. 37, 27, pl. 10, figs. 5-8, 1887; Knowlton, U. S. Geol. Surv. Prof. Paper 101, 259, pl. 53, fig. 5, 1917. Dryophyllum Jalcalum Ward, U. S. Geol. Surv., 6th Ann. R«pt., 551, pl. 37, fig. 10, 1886; idem, BuU. 37, 27, pl. 11, fig. 1, 1887; KJiowlton, U. S. Geol. Surv. Bull. 163, 42, pl. 8, fig. 1, 1900. Quercus gracilis Newberry, U. S. Nat. Mus. Proc, vol. 5, 504, 1883; U. S. Geol. Surv. Mon. 35, 75, pl. 67, fig. 4, 1898. Quercus doljensis Pilar. Ward, U. S. Geol. Surv., 6th Ann. Rept., 551, pl. 36, figs. 9, 10, 1886; idem, Bull. 37, 25, pl. 9, figs. 4, 5, 1887. Quercus whitmani Knowlton, U. S. Geol. Surv. Prof. Paper 155, 52, pl. 17, fig. 5, 1930. Quercus turhulenla HoUick, U. S. Geol. Surv. Prof. Paper 159, 70, pl. 38, fig. 3, 1930. Rhus pseudo-meriani Lesquereux, Rept. U. S. Geol. Surv. Terr., vol. 7, 293, pl. 58, fig. 11, 1878. Fraxinus? princetoniana Knowlton, U. S. Geol. Surv. Prof. Paper 130, 161, pl. 22, fig. 7, 1922. Thi.s type of leaf is recorded from 5 of the 6 Medicine Bow localities and is usually one of the dominants. The large number of specimens collected for study were at first considered to repre- sent 5 distinct species on the basis of identity with figured specimens. It was confusing, however, to discover that there were intermediate types present with a few characteristics of one species but the remaining characteristics of another species. After consulting the type collections at the U. S. National Museum, it has seemed necessary to regard all of these species as variants of a single rather well-defined leaf type. The original specimen of Dryophyllum {Quercus) subfalcatum was coUected from what is now called the Mesaverde formation at Point of Rocks, Wyoming. In the following decade Ward and Newberry independently described additional specimens of this same type of leaf from the same locality, as Dryophyllum bruneri Ward and Quercus gracilis Newberry. Neitlier author mentioned the work of the other nor made any reference to tlie similar leaf described by Lesque- reux. Knowlton clarified this situation by synonymizing both species with Dryophyllum sub- falcatum,^ although D. bruneri was inadvertently restored to its specific status in Bulletin 696. From the shghtly younger Black Buttes flora of the same region specimens of essentially identical type were variously described as D. aquamarum Ward, Quercus doljensis Pilar, and Rhv^ pseudo-meriana Lesquereux. In a recent synonymy Brown has included the first of these in Dryophyllum subfalcatum.'^ I ani including the last two, already combined as one species by Knowlton,^ as possessing no recognizable differences. Brown has also inchided, on justifiable evidence, the specimens hitherto referred to Dryophyllum falcatum Ward, Querciis whitmani Knowlton, and Q. turbulenta Hollick from scattered beds of late Cretaceous age. To these I would add the specimen of Fraxinusf princetoniana from the Laramie flora. Among my specimens which I have collected from the lower Lance east of Lance Creek, Wyoming, there are several specimens of Dryophyllum subfalcatum to add to the Lance occurrences recorded by Knowlton ^ and Brown.^ Moreover, a single specimen of the same species was tenta- tively described as Juglans? confusa by Knowlton in an unfinished manuscript on a collection from the Lance formation of "Converse County," Wyoming. As now conceived by Dr. Brown and me, this is a valuable index species of the late Creta- ceous, being well defined, widespread, and usually abundant. It is unfortunate that as good use cannot be made of it in ecologic and climatic considerations owing to its uncertain botanical affinities. Cretaceous leaves of tliis type are usually referred without further consideration to the genus Dryophyllum, held to be prototypic of the family Fagacese. Tertiary specimens of the same general character are ordinarily referred to Castanea or Quercus, though often on insufficient evi- dence. I would hesitate, in the absence of seeds, to remove the present species to either Castanea or Quercus, although the resemblance to many species of these genera is admittedly striking. Occurrence — Corson Ranch, Wyoming, Locs. P. 371, P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374; Craig, Colorado, Loc. P. 376. Collection—V. C. Mus. Pah, Plesiotypes Nos. 1313, 1314, 1315. > Knowlton, F. H., U. S. Geol. Surv. Bull. 163, 41, 1900; idem, BuU. 690, 246, 1919. 2 Brown, R. W., U. S. Geol. Surv. Prof. Paper, manuscript, 1937. ' Knowlton, F. H., U. S. Geol. Surv. BuU. 696, 525, 1919. ' Knowlton, F. H., Wash. Acad. Sci. Proc, vol. 11, 211, 1909. ' Brown, R. W., op. cit. FOX HILLS AND LOWEK MEDICINE BOW 53 Genus QUERCUS Linnd Quercus viburnifolia Lesquereux (Plate 6, Figs. 3, 5, 7) Quercm mbumifolia Lesquereux, Rept. U. S. Geol. Surv. Terr., vol. 7, 1.59, pl. 20, fiK.s. 11, 12, 1878; U. S. Gcol. and Geog. Surv. Terr., Ann. Rept. (1876), 50.5, 1878; Harvard Coll. Mus. Comp. Zoology Bull., vol. 16, 46, 1888; Knowlton, Washington Acad. Sci. Proc, vol. 11, 191, 207, 1909; U. S. Geol. Surv. Prof. Paper 130, 127, 1922; idem, Prof. Paper 155, .50, pl. 16, figs. 3-8, pl. 17, fig. 1, 1930. Alnus auraria Knowlton and Cockerell. Knowlton, U. S. Geol. Surv. Prof. Paper 155, 49, pl. 15, fig. 6, 1930. Beiula fallai Lesquereux (in part). Knowlton, ibid., 50, pl. 15, figs. 7-9, pl. 16, figs. 1, 2. Celastrus gaudini Lesquereux (in part). Knowlton, ibid., 99, pl. 45, fig. 9 (not figs. 3, 4). Fraxinus sp. Knowlton, ibid., 124, pl. 58, fig. 7. Populus denverensis Knowlton, ibid., 61, pl. 23, fig. 5. Quercus purdonensis Knowlton, ibid., 52, pl. 17, fig. 2. Quercus whHei Lesquereux. Knowlton, ibid., 53, pl. 17, fig. 3, pl. 18, fig. 1. Vibumum coniorlum Lesquereux?. Kjiowlton, ibid., 127, pl. 55, fig. 1. The long synonymy cited above illustrates clearly the prevalence of this type of leaf in the Lance, Laramie, Black Buttes, Dawson, and Denver floras. The large number of "speoies" from the Denver flora which arc here combined with Quercus viburnifolia might be open to criticism. After studying the type specimens at the U. S. National Museum I find it impossible, however, to segregate them into distinct species with truly recognizable differences. The alternatives were either to merge them into a single species, whose leaves are somewhat variable — though no more so than those of a single modern tree; or to separate the specimens into about 10 species based on minor differences in shape, venation, or marginal characters. To avoid rather than produce con- fusion, it was thought best to combine the specimens, incidentally adding greater value to the species for correlation purposes. Specimens of this species were collected from 5 of the Medicine Bow localities. In none of these is the species as abundantly represented as it is in the collections from the Denver formation. Its presence in both the Laramie and Lance formations is reported in Professional Paper 130 cited above, as well as in an unpubhshed manuscript of Knowlton's seen at the U. S. National Museum. It is apparent that this species is not present in any of the extensive Fort Union or Eocene col- lections at the Museum, nor has it ever been reported in beds either older or younger than post- Montana Cretaceous. The generic reference of this species to Quercus seems open to question. I have coasulted herbarium material of all genera of the Fagaceae and have not succeeded in finding any leaves of comparable form. The general shape, venation, and character of marginal teeth of the fossil leaves seem rather to relate them to the Betulaceae or the CaprifoHace£e. Occurrence — Corson Ranch, Wyoming, Loc. P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374; Craig, (jolorado, Locs. P. 375, P. 376. Collection—\J. C. Mus. Pal., Plesiotypes Nos. 1316, 1317, 1318. Family MORACEiE Genus FICUS (Tournefort) Linn6 Ficus planicostata Lesquereux (Plate 5, Figs. 3, 4, 5, 7) Ficus planicostata Lesquereux, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1872), 393, 1873; Rept. U. S. Geol. Surv. Terr., vol. 7, 201, pl. 31, figs. 1-8, 10-12, 1878; Knowlton, U. S. Geol. Surv. BuU. 163, 52, pl. 10, fig. 4, pl. 12, figs. 2, 3 (not 4), 1900; Washington Acad. Sci. Proc, vol. 11, 211, 1909; Cockerell, Univ. Colorado Studies, vol. 7, 151, 1910; Ivnowlton, U. S. Geol. Surv. Prof. Paper 130, 131, 1922; idem, Prof. Paper 134, 82, pl. 9, fig. 2, 1924. Ficus planicostata goldiarM Lesquereux, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1873), 399, 1874; Rept. U. S. Geol. Surv. Terr., vol. 7, 202, pl. 33, figs. 1-3, 1878; Ivnowlton, U. S. Geol. Surv. Prof. Paper 155, 70, pl. 28, fig. 5, 1930. Ficus planicostata cUnioni (Lesquereux) Knowlton, U. S. Geol. Surv. Bull. 152, 103, 1898; idem, Prof. Paper 101, 303, pl. 76, fig. 3, 1917. Ficus clinioni Lesquereux, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1872), 393, 1873. Ficus planicostata magnifolia Ivnowlton, U. S. Geol. Surv. Prof. Paper 130, 133, pl. 10, fig. 3, 1922. Ficus leei Knowlton (in part), U. S. Geol. Surv. Prof. Paper 101, 261, pl. 39, figs. 2-5 only, 1917; idem, Prof. Paper 98, 338, pl. 90, fig. 2, 1917. 54 CONTRIBUTIONS TO PAL^ONTOLOGY Ficus pseudopopulxis Lesquereux (in part), Knowlton, U. S. Geol. Surv. Prof. Paper 101, 304, pl. 72, figs. 3, 4 only, 1917; idem, Prof. Paper 134, 83, pl. 7, fig. 4, pl. 9, fig. 3, 1924; idem, Prof. Paper 155, 66, only pl. 25, figs. 3-5, 1930. Ficus neoplanicostata Ivnowlton (in part), U. S. Geol. Surv. Prof. Paper 101, 303, pl. 73, fig. 4, pl. 74, figs. 2, 3, pl. 76, fig. 4, 1917; idem, Prof. Paper 134, 82, pl. 9, fig. 4, 1924; idem, Prof. Paper 155, 69, only pl. 28, figs. 3-7, 1930. Ficus prcelatifolia Ivnowlton, U. S. Geol. Surv. Prof. Paper 98, 338, pl. 87, fig. 4, 1917. Pterospermites neomexicanus Knowlton, ibid., 341, pl. 90, fig. 6. Ficus impressa Knowlton, U. S. Geol. Surv. Prof. Paper 130, 134, pl. 7, figs. 1-3, pl. 16, fig. 3, 1922. The specimens from Black Buttes, Wyoming were originally described by Lesquereux as f ollows : "Leaves of medium size, subcoriaceous, entire, elliptical or broadly oval, .slightly acuminate or obtuse, rounded to a short, thick, petiole, palmately three-nerved from the top of tbe petiole, rarely from a short distance above the base; primary and secondary nerves broad, fiat, all campto- drome, as well as their divisions." Leaves of precisely this description are by f ar the most abundant of all species in the Medicine Bow collections. Over 50 specimens of complete or nearly complete leaves were collected for study. In arranging this suite for identification it was apparent that there was a perfect series from small ones 1.2 by 2 cm. in dimensions to very large ones 7 by 12 cm. It was also found possi- ble to match individual specimens perfectly with the figures and type specimens of each of the species listed in the above synonymy. These various species had been distinguished from Ficus planicostata by minor details, chiefly of venation, but were in each case discussed as difficult to separate from that species. They are here regarded as mere variants of the average leaf form. In 1923 Berry discussed the difficulties involved in the taxonomic separation of leaves of this type and concluded that Ficus planicostata and its varieties were be.st considered conspecific with Ficus mississippiensis (Lesquereux) Berry.' In comparisons of my large suite with all available specimens of both these species and others at the U. S. National Museum I have found, however, that the following criteria can serve to distinguish them: Ficus planicostata 1. Size: Average less than 10 cm. long. 2. Shape: Always elliptical or broadly oval, widest near the middle. 3. Tip: Always obtuse and abruptly narrowed to a blunt point. 4. Venation: Lateral primaries basilar. Ficus mississippiensis 1. Size: Average more than 10 cm. long. 2. Shape: Always ovate or ovate-lanceolate, widest well below the niiddle. 3. Tip; Always acute and graduaUy narrowed from the middle of the leaf to a long, slender point. 4. Venation: Lateral primaries suprabasilar. It must be admitted, of course, that it is largely a matter of taxonomic judgment as to which characters of leaves are more or less important in distinguishing one species from another. In this case, however, I feel that my conclusions were inevitable because of the constancy with which the listed characters were maintained in so large a collection of leaves of all sizes. As here conceived, Ficus planicostata is a valuable index of late Cretaceous deposits in the Rocky Mountain region, though it may possibly range into the early Paleocene. It makes its first appearance in the upper Mesaverde formation of Wyoming, is well represented in the Vermejo, Raton, Fruitland, and Animas formations, and is abundant in the Laramie, Denver, and Black Buttes floras. Although it has been reported from only one locality in the Lance,^ I have seen other specimens of it from the Lance of "Conver.se County," Wyoming, at the U. S. National Museum, which Knowlton referred to Phyllites ficifolia, new species.' The species has never been found in beds of indisputable Paleocene age, although collections from such beds, namely the Fort Union formation, are very extensive. Related descendants of this species, such as Ficus inississip- ' Berry, E. W., U. S. Geol. Surv. Prof. Paper 131-A, 9, pls. 6-8, 1923. «Knowlton, F. H., Washington Acad. Sci. Proc, vol. 11, 211, 1909. ' Knowlton, F. H., unpublished manuscript. FOX HILLS AND LOWER MEDICINE BOW 55 piensis and Ficus plincrva Chanoy and Sanborn ' appcar to havo survived into the Tertiary on the Gulf coastal plain and the Pacific Coast respectively. Among living forms I have noted a general similarity of the ieaves of Ficus planicostata to a number of genera: Cocculus (Cebatha) and Ilypcrbiena of the MenispermaceJE, Alchornea and Mallotus of tlie Euphorbiaceic, Pterosperrnum (Sterculiacepe), Ilampea (Bonbacacese), Colubrina (Rhamnacea;), and Ficus (Moracea;). After consulting all available herbarium material of both eastern and western liemisphere representatives of thesc genera at The New Yorlc Botanical Garden, I am convinced that tlic refcrence to Ficus is not only reasonablo but probably correct. A close resemblance vvas noted between the fossil specimens and the leaves of Ficus colubrincB Standley ^ of Guatemala and Ficus integrifolia Elm.' of the Philippines. Except for their re- motely toothed margins the lcaves of Ficus riedelii Miq.'' of Celebes are also comparable. Occurrence — Corson Ranch, Wyoraing, Locs. P. 371, P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374; Craig, Colorado, Locs. P. 375, P. 376. Collection—V . C. Mus. Pal., Plesiotypes Nos. 1391, 1319, 1320, 1321. Ficus cockerelli Knowlton (Plate 7, Fig. 2) Ficus cockerelli Knowlton, U. S. Geol. Surv. BuII. 696, 273, 1919; idem, Prof. Paper 130, 132, pl. 12, fig. 2, pl. 23, figs. 1, 2, 1922. Ficus lalifolia (Lesquereux) Knowlton, U. S. Geol. Surv. Bull. 152, 102, 1898; idem, Prof. Paper 101, 304, 1917. (Homonym, Kunth, 1846.) Ficus planicostaia lalifolia Lesquereux, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1872), 393, 1873; Rept. U. S. Geol. Surv. Terr., vol. 7, 202, pl. 31, fig. 9, 1878. There are about 10 specimens of this type, which are similar except in their slightly smaller size to the figured and type specimens of this species. I was at first inclined to regard these specimens as variants of Ficus planicostata. Among the large number of specimens of the latter species, however, there are none which vary in the direction of the round, broad leaves of F. cockerelli. The heart-shaped base and the venation of this species are also distinctive. There is a resemblance to F. speciosissima Ward from the Mesaverde formation * and F. leei Knowlton from the Verraejo,^ which I believe are conspecific with each other. These, however, are decidedly more heart-shaped at the base, are longer in proportion to their width, and possess notable dif- ferences in both secondary and tertiary venation. Ficus cockerelli is reported in the above citations to be questionably present in the Raton fiora and fairly common in both the Laramie and Black Buttes floras. It is said to occur also in the Wilcox of the Gulf states.' I have not seen the Wilcox specimens, which Berry more recently has considered synonymous with Ficus mississippierisis.^ Among modern leaves these specimens are in my opinion more similar to the leaves of Cocculus, Cissampelos, and Menisperminn (Menispermacese) than to any species of Ficus which I have seen. Except for their thinner texture, it is difficult to distinguish the fossil speciraens from the leaves of Cocculus {Ccbatha) carolinus (L.) DC^ Occurrence — Corson Ranch, Wyoming, Locs. P. 371, P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374. Colleciion—lJ. C. Mus. Pal., Plesiotype No. 1322. > Chaney, R. W., and Sanborn. E. I., Carnegie Inst. Wash. Pub. No. 439. 66, 1933. ' See sheet No. 7668 at The New Vork Botanieal Garden. ' See sheet No. 76244. < Chaney, R. W., and Sanborn, E. I., Carnegie Inst. Wash. Pub. No. 439, 67, 1933. ' Ward, L. F., U. S. Geol. Surv. Bull. 37, 39, pl. 21, fig. 3, 1887. • Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 101, 261, pl. 39, figs. 1, 6, pl. 40, figs. 1. 2, 1917. ' Berry. E. W., U. S. Geol. Surv. Prof. Paper 91, 199, 1916. ' Berry, E. W., U. S. Geol. Surv. Prof. Paper I3I-A, 9, 1923. ' See sheet No. 18, Alabama, in the Princeton University herbarium. 56 CONTRIBUTIONS TO PAL^ONTOLOGY Ficus trinervis Knowlton (Plate 6, Figs. 1, 4) Ficus trinervis ICnowlton, U. S. Geol. Surv. Bull. 163, 42, 1900; Cockerell, Am. Mus. Nat. History Bull., vol. 24, 89, 1908; Knowlton, Washington Acad. Sci. Proc, vol. 11, 197, 1909. Cinnamomum affine Lesquereux (in part), Rept. U. S. Geol. Surv. Terr., vol. 7, 219, pl. 37, fig. 5 only, 1878; Ward, U. S. Geol. Surv., 6th Ann. Rept., 553, pl. 67, figs. 1-3, 1886; idem, BuII. 37, 50, pl. 24, figs. 3-5, 1887. Ficus prcptrinervis Knowlton (in part), U. S. Geol. Surv. Prof. Paper 101, 263, 304, pl. 41, figs. 1-3 only, pl. 42, fig. 1, 1917; idem, Prof. Paper 98, 338, 1916; idem, Prof. Paper 155, 71, pl. 28, fig. 8 only, 1930. Ficus haddeni Knowlton, U. S. Geol. Surv. Prof. Paper 101, 260, pl. 38, figs. 6, 7, 1917. Malapoenna louisvilknsis Ivnowlton, U. S. Geol. Surv. Prof. Paper 130, 144, pl. 7, fig. 5, 1922. Ficus moplanicosta Knowlton (in part), U. S. Geol. Surv. Prof. Paper 155, 69, only pl. 29, figs. 3, 4, 1930. Cinnamomum sp. Ivnowlton, ibid., 87, pl. 39, fig. 1. This type of leaf is not abundant, being represented by only 6 specimens from 3 localities. None of these is perfectly complete, but enough is preserved to show that these can not be sepa- rated from the type specimens (not figured by Knowlton) of Ficiis trinervis, and the figured and type specimens erroneously referred to Cinnamojnum affine which Knowlton properly referred to Ficus trinervis. Knowlton justifiably distinguished these specimens from Cinnamomum affine by the character of the basal secondaries, which "pass to extreme base of the blade and arise with and at the same point as the midrib." ^ In his report on the Vermejo flora Knowlton subsequently described what he called a very closely related species, Ficus pra:trinervis, which he di.stinguished from F. trinervis by a slight outward arching of the basal secondaries in the upper half of the blade. Since this minor difference does not appear to be consistent in my specimens or in the type speci- mens or duphcates at the U. S. National Museum, I feel justified in uniting these two species. The characters which appear to be diagnostic of Ficus trinervis are as follows: shape broadly ovate to ovate-lanceolate; obtuse to slightly acuminate tip, and abruptly rounded decurrent base; tri-nerved from below the actual base of the blade, so that the basal laterals form the margin of the leaf down to the petiole; margin entire; lateral secondaries given off from the midrib at very acute angles, almost paralleUng the niidrib and finally joining the next secondaries above by marginal loops. The four species Usted above as synonymous with F. trinervis all possess this same set of characters. As now known, Ficus trinervis is present in the Medicine Bow, Mesaverde, Vermejo, Fruit- land, Raton, Laramie, Lance, Denver, Dawson, and Black Buttes floras. To my knowledge it has never been reported in beds of indisputable Paleocene or later age. The generic reference of this species to Ficus has in no way been substantiated by my survey of all species of this genus in the herbarium of The New York Botanical Garden. The general shape, margin, and venation of the fossil leaves are, in my opinion, more likely menispermaceous, yet I have been unsuccessful in finding a single species of this family in which the leaves are basally decurrent. Occurrence — Corson Ranch, Wyoming, Locs. P. 371, P. 372; Craig, Colorado, Loc. P. 376. Collection—V. C. Mus. Pal., Plesiotypes Nos. 1323, 1324. Ficus crossii Ward (Plate 7, Fig. 3) Ficus crossii Ward, U. S. Geol. Surv., 6th Ann. Rept., 552, pl. 44, fig. 7, 1886; idem, Bull. 37, 39, pl. 21, fig. 2, 1887; {not Knowlton, idem, Prof. Paper 130, pl. 11, fig. 2, which = Rhamnus goldianus). Cornus ernmonsii Ward (in part), U. S. Geol. Surv., 6th Ann. Rept., 553, pl. 47, fig. 3 only, 1886; idem, BuU. 37, 55, pl. 26, fig. 3 only, 1887. Cornus impressa Lesquereux. Knowlton, U. S. Geol. Surv. Bull. 163, 68, 1900. Cornus pradmpressa Knowlton, U. S. Geol. Surv. Prof. Paper 130, 159, pl. 14, fig. 5, pl. 19, fig. 2a, 1922; Brown, U. S. Geol. Surv. Prof. Paper, manuscript, 1937. Rhamnites knowlloni Berry (in part), U. S. Geol. Surv. Prof. Paper 185-F, 131, pl. 26, fig. 1, 1934. Specimens of this type of leaf are not common but are usually rather well prcserved. The specimen here figured is virtually complete in all details, which are similar to the specimens referred to the 5 distinct species above. It should be pointed out that the figures of Ficus crossii and Cornus emmon.sii were not correctly drawn nor properly restored, as has been previously noted by 1 Knowlton, F. H., U. S. Geol. Surv. BuU. 163, 43, 1900. FOX HILLS AND LOWER MEDICINE BOW 57 Knowlton.' The tj'pe spccimens of both of thcse spccies are not distinguishable in any way from my specimens nor from the specimen of Cornus prmmpressa. I have not seen the actual specimen of Rhamnites knowltoni from the Lance formation, but the figure seems quite unlike the type specimens of this species from the Eocene of the Gulf Coast,^ and is not distinguishable from the specimens here referred to Ficus crossii. Such a reference is substantiated by the presence of several indistinguishable leaves in a coUection which I recently made from the lower Lance near Ekalaka, Alontana. In view of the discovery of more complete specimens than hitherto known, it seems desirable to ampHfy the descriptions previously given of this species. Descripiion {Supplementary) — Leaves elHptic to oval, with a broadly cuneate base and abruptly short-acuminate tip; lcngth 8 to 10 cm.; width 5 to 6cm.; midrib thick in depth, especially below, straight or very slightly curved; secondaries about 10-12 pairs, mainly alternate, at an angle of about 60° near the base to as low as 35° near the apex, slightly curved upward, unbranched, camptodrome, arching upward just inside the margin; tertiary veins fairly strong, mainly un- broken, at angles of 90° to thc midrib ncar the apex to as low as 65° to the midrib in the lower half ; finer veins forming an irregular meshwork; margin entire; texture firm. This species as now known is present in the Mesaverde, Colgate, Laramie, lower Lance, and lower Medicine Bow formations, with the closely related Rhamnites knowltoni occurring in the Eocene of the Gulf Coast. There do not appear to be any Fort Union species with which conf usion might result. Among living forms there appear to be very few species of Ficus with leaves of this general character. I have surveyed the foUowing genera at The New York Botanical Garden as alterna- tive possibilities for comparison: Anona (Anonacese), Cornus and Nyssa (Cornacese), Ficus (Moraceae), Magnolia (Magnohaceffi), Rhaninus (Rhamnaceae), Cephalanthos (Rubiacese), and Diospyros (Ebenaceee). The cIo.sest resemblance among these genera was with Magnolia, particularly M. virginiana L. (= M. glauca L.) and the smaller leaves of M. grandiflora L. Occurrence — Corson Ranch, Wyoming, Loc. P. 372; Elk Mountain road, Wyoming, Loc. P. 374; Craig, Colorado, Locs. P. 375, P. 376. Collection—ll. C. Mus. Pal., Plesiotype No. 1325. Ficus tesselata Lesquereux (Plate 6, Fig. 2) Ficus tesselata Lesquereux (MS.), Knowlton, U. S. Geol. Surv. Prof. Paper 101, 266, pl. 41, fig. 5, 1917. The collections contain only 2 fairly complete specimens and several fragments of this type of leaf . These are essentially similar to the figured and type specimen of Ficus tesselata from the Vermejo formation. Knowlton has pointed out the resemblance to Rhaninus salicifolius Les- quereux,' from which it differs decidedly in its shape, its fewer, more acute secondaries, and its open tertiary venation. It seems even closer to Laurus lanceolata Knowlton from the Laramie and Dawson formations.^ The known specimens of this species, unfortunately, are too few and too fragmentary to make possible a detailed comparison. Knowlton has recorded Ficus tesselata also from the Lance formation in an unpublished manuscript at the U. S. National Museum. I can see httle justification for the reference of this species to the genus Ficus. In a complete survey of both eastern and western hemisphere species of Ficus at The New York Botanical Garden I was unable to find any leaf forms with comparable characters. The family Lauraceae, on the other hand, contains many comparable leaf forms, chiefly in the genus Nectandra (cf. N. glauca ^ and N. cuspidata * of South America). Occurrence — Corson Ranch, Wyoming, Loc. P. 371; Elk Mountain road, Wyoming, Loc. P. 374. Collection—\]. C. Mus. Pal., Plesiotype No. 1326. ' Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 130, 139. 160, 1922. ' Berry, E. W.. U. S. Geol. Surv. Prof. Paper 131, 16, pl. 12, fig. 7, 1923. 3 Lesquereux, Leo, Rept. U. S. Geol. Surv. Terr., vol. 7, 282, pl. 53, figs. 9, 10, 1878; Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 130, 154, pl. 15, fig. 4, pl. 19, fig. 2b, 1922. » Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 130, 143, pl. 21, fig. 7, 1922; idem, Prof. Paper 155, 85, pl. 38, fig. 9, 1930. ' See sheet No. 708, Boli^na, at the Princeton University herbarium. ' See sheet No. 707, Mapiri, Bolivia. 58 CONTRIBUTIONS TO PAL^ONTOLOGY Ficus coloradensis Cockerell (Plate 6, Fig,. 6) Ficus coloradensis Cockerell, Torreya, vol. 10, 223, 1910; Knowlton, U. S. Geol. Surv. Prof. Paper 130, 134, pl. 22, fig. 1, 1922; idem, Prof. Paper 155, 69, 1930. Ficua irregularis (Lesquereux) Lesquereux (in part), R/ipt. U. S. Geol. Surv. Terr., vol. 7, 196, pl. 34, figs. 4-7, 1878; Harvard CoU. Mus. Comp. Zoology Bull., vol. 16, 50, 1888; Ward, U. S. Geol. Surv., 6th Ann. Rept., 552, pl. 44, fig. 4 only, 1886; idem, Bull. 37, 38, pl. 20, fig. 4 only, 1887; Ivnowlton, U. S. Geol. Surv. BuII. 163, 51, 1900. Ulmusf irregularis Lesquereux, U. S. Geol. Surv. Terr., 6th Ann. Rept. (1872), 378, 1873. Only 4 specimens in the coUections are of this type. None is completely preserved, though enough is shown to indicate .similarity to the figured and type specimens of this species. There is httle I can add to Ivnowlton's able discussion of its status.^ It is present in the Laramie, Denver, and Medicine Bow formations, and occurs doubtfully in the Mesaverde, Hanna ("Upper Laramie" of Knowlton), and Black Buttes fioras. I can offer no suggestions regarding the botanical affinities of this leaf form. Occurrence — Elk Mountain road, Wyoming, Loc. P. 374. Collection—U. C. Mus. Pal., Plesiotype No. 1327. Ficus cowanensis Knowlton (Plate 7, Fig. 1) Ficus cowanensis Knowlton, U. S. Geol. Surv. Prof. Paper 130, 139, pl. 8, fig. 6, pl. 9, figs. 2, 3, 1922. There are several incomplete specimens in the collections which are quite distinct in their large size, entire margin, obovate-Ianceolate shape, extremely thick midrib, and thin, widely spaced secondaries. They agree perfectly with the description and type specimens of Ficus cowanensis from the Laramie formation. This species has never been found completely preserved. I suspect that it may ultimately be found to be the same as some of the leaves usually referred to Magnolia or Laurus, as for exam- ple, Magnolia leei Knowlton, M. hilgardiana Lesquereux, or Laurusf caudata Knowlton.^ Occurrence — Elk Mountain road, Wyoming, Loc. P. 374. Collection—l]. C. Mus. Pal., No. 1328. Family PLATANACE^ Genus CREDNERIA Zenker Credneria protophylloides Knowlton (Plate 8) Credneria protophylloides Knowlton, U. S. Geol. Surv. Prof. Paper 101, 267, pl. 46, 1917. There is a single specimen and its counterpart and many incomplete specimens from Locality P. 375 which are similar in size and essential details to the type and figured specimens of this species from the Vermejo formation. The Medicine Bow specimen is somewhat larger than the figured Vermejo specimen, but is identical with the large, unfigured specimen described by Knowl- ton from the same formation. There is nothing to be added at the present time to the accurate description and discussion of this species by Knowlton. He has already adequately pointed out the superficial resemblance of this species to several American species of Protophyllum, and its greater similarity to a number of European members of this genus. Comparable types appear to be rare in the late Cretaceous and early Tertiary of the Rocky Mountains. Hollick has described a number of species pf Credneria and of the related genera Protophyllum, Pseudoprotophyllum, and Pseudoaspidiophyllum from the Cretaceous of AJ.aska.^ Most of these are represented by large leaves with the same ■ Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 130, 134, 1922. = Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 101, 310, 313, 316, pl. Sl, fig. 2, pl. 85, fig. 1, pl. 89, fig. 1, 1917. " Hollick, A., U. S. Geol. Surv. Prof. Paper 150, 86-97, 1930. FOX HILLS AND LOWER MEDICINE BOW 59 distinctive vonation as in Crcdncria protophylloides. With the possible exception of C. intermedia,'- which is unfortunatcly not coaipletcly known, none of the Alaskan forms are entire-margined and they are therefore not apparently conspecific with the Mcdicine Bow and Vermejo species. Among living forms, the widcsprcad Cretaceous genus Credneria has apparently no closely similar leaf forms. Venation characters and general shape and size seem to indicate aflinities with Platanus.^ Occurrence — Craig, Colorado, Loc. P. 375. Collection—\] . C. Mus. Pal., Plesiotype No. 1329. Family NYCTAGINACE^ Genus PISONIA (Plumier) Linn6 Pisonia? racemosa Lesquereux (Plate 9, Fig. 2) Pisonia racemosa Lesquereux, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1873), 400, 1874; Rept. U. S. Geol. Surv. Terr., vol. 7, 209, pl. 35, fig. 4, 1878. The small, delicate cluster of seed-like bodies here figured is not distinguishable from the specimens of this species from Black Buttes, Wyoming. Their exact nature is problematical and their reference to Pisonia is by no means definite, as Lesquereux clearly stated in his original analysis. If they are seeds they are considerably smallcr and wider at the base than any seeds of Pisonia (= Torrubia Vellozo) which I have been able to find for comparison. More and better material will have to be obtained to make possible a definite systematic assignment. Remains of this type are evidently rarely preserved, as they have not been recognized in deposits other than those at the type locality and those of the Medicine Bow formation here recorded. Occunence — North of Walcott, Wyoming, Loc. P. 373. Collection—U. C. Mus. Pal., Plesiotype No. 1330. Family LAURACE.E Genus CINNAMOMUM (Tournefort) Linn6 Cinnamomum linifolium ICnowIton (Plate 9, Fig. 1) Cinnamomum linifolium Knowlton, U. S. Geol. Surv. Prof. Paper 101, 319, pl. 88, figs. 3-7, 1917; idem, Prof. Paper 155, 86, pl. 59, fig. 3, 1930. This type of leaf is rare in the collections, being represented at two localities by only three fragments, the best of which is here figured. It agrees in essentially all observable characters with the description and type specimens of this species from the lower Raton and Dawson formations of Colorado. Knowlton has previously pointed out tlie general resemblance of this species to Cinnamomum ohlongatum Berry and C. vera Berry from the Gulf Coast Eocene.^ To my knowledge there are no other fossil forms known in North America with which even a general comparison can be made. Knowlton was originally in doubt concerning the generic reference of this species. There is unquestionably a resemblance to Cinnamomum. I have also found leaves of essentially the same type in herbarium specimens of Cocculus larvifolius D.C., and Lindera caudata Benth., of Asia. More complete specimens will be necessary, however, before detailed comparisons can be made. Occurrence — North of Walcott, Wyoming, Loc. P. 373. Collection—V . S. Nat. Mus. Coll., Plesiotype No. 1331. ' HoUick, A., ibid., 89, pl. 55, fig. 1. «Seward. A. C, Plant Life through the .\gea, 404, 1933; Potonie-Gothan. Lehrbuch der Palaobotanik, 377, 192L ' Berry, E. W., U. S. Geol. Surv. Prof. Paper 91, 297, pl. 79, figs. 1-8, pl. 83, fig. 6, pl. 87, fig. 4, 1916. 60 CONTRIBUTIONS TO PAL^ONTOLOGY Cinnamomum affine Lesquereux (Plate 9, Figs. 3, 4) Cinnamomum affine Lesquereux (in part), Rept. U. S. Geol. Surv. Terr., vol. 7, 219, pl. 37, figs. 1-4, 7 only, 1878; Knowlton, U. S. Geol. Surv. Prof. Paper 130, 145, pl. 8, fig. 4, pl. 17, fig. 6, 1922. The nearly complete leaf (figureS) is the best of about a dozen of this type in the coUections. It is not wholly typical. It has the tri-nerved venation chverging from a point along the midrib well above the petiole, which is so typical of Cinnamomum affine, yet it differs slightly in shape, which is wider and less pointed than the average of oiir specimens (figure 4), and in the presence of an indistinct pair of veins below the lateral primaries. These veins are not usually present in the recorded leaves of this species, though occasionally seen, as in Lesquereux's figure 7. Judging from the specimens in the Medicine Bow collections, these veins are more common than has hitherto been observed. It may be that this character is important enough to distinguish these leaves as a separate species. For the present I consider them conspecific. This species is known at present only from the Laramie, Colgate, and Medicine Bow forma- tions. It may possibly also occur in the Denver formation in what has been referred to Cinnamo- mum sp., for want of complete specimens.' A larger suite of Laramie specimens might also indicate that C. laramiensis Knowlton ^ is but a narrow variant of the normal C. affine type. A resemblance of these fossil specimens to the leaves of the modern Cinnamomum cannot be denied. The subsidiary basal veins and the faintness of the tertiary venation, however, are features not observed in Cinnamomum. Occurrence — Corson Ranch, Wyoming, Loc. P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374. Colledion—U. C. Mus. Pal., Plesiotypes Nos. 1332, 1333. Genus LAUROPHYLLUM Goppert Laurophyllum meeki Dorf, n. sp. (Plate 7, Fig. 5) Diospyros? ficoidea Lesquereux (in part), Rept. U. S. Geol. Surv. Terr., vol. 7, 231, pl. 40, fig. 6 only, 1878. Daphnogene elegans Watelet. Ward, U. S. Geol. Surv., 6th Ann. Rept., 553, pl. 47, fig. 4, 1886; idem, Bull. 37, 51, pl. 35, fig. 1, 1887. There are 2 specimens which are indistinguishable from those previously referred to the above two species of Lesquereux and Ward. Their specimens were both collected from the same locality, Black Buttes, Wyoming, and are clearly the same species. The specimen here figured from the Medicine Bow formation looks, indeed, like a counterpart to the specimen figured by Ward, which is clearly not, as he himself intimates, referable to Watelefs European species. The generic reference to Lauro-phyllum is purposely non-committal, though indicating resemblance to lauraceous leaves. It is unfortunate that in all known specimens of this species the base and apex are both destroyed, making a complete description impossible. Description — Leaves lanceolate to elhptic, narrowed gradually to the base and apex, whose specific characters are not known; length apparently about 12 to 15 cm.; width 4 to 5 cm. at the widest portion just below the middle; midrib fairly thick, straight; secondaries prominent, 6 to 8 subopposite pairs, branching acutely from the midrib at angles of 25° to 30°, curving shghtly upward, converging gradually upon the leaf margins in a long series of prominent, quadrate loops; basal secondaries with prominent looping tertiaries on their lower side; remaining tertiary veins prominent, regular, and coar.se, running at angles close to 90° to the midrib; margin entire, slightly undulate; texture subcoriaceous. There do not appear to be any other described fossil forms of this genus, or related genera, with which this species can be compared. I have named the species for Professor F. B. Meek, who collected the first specimen of this type at Black Buttes, Wyoming. Occurrence — Elk Mountain road, Wyoming, Loc. P. 374; Craig, Colorado, Loc. P. 376. Collection—U. C. Mus. Pal., Plesiotype No. 1334. ' Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 155, 87, pl. 39, fig. 1, 1930. ' Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 130, 146, pl. 22, fig. 3. 1922. FOX HILLS AXD LOWER MEDICINE BOW 61 Genus LAURUS (Toumefort) Linnd Laurus socialis Lesquereux? (Plate 7, Fig. 4) Laurus socialis Lesquereux, Rept. U. S. Geol. Surv. Terr., vol. 7, 213, pl. 36, figs. 1-4, 7, 1878; Knowlton, U. S. Geol. Surv. Prof. Paper 101, 317, pl. 91, fig. 5, 1917; idem, Prof. Paper 155, 85, pl. 38, fig. 4, 1930. There are only 3 fragmentary spccimens of this typc in the collections. They are referred with doubt to this species. The spccimens are closely similar to those from the Raton and Dcnver formations, which seem to mc, however, to differ considcrably from the original type specimens from Evanston, Wyoming. In these the shape is more lanccolatc and the secondaries are more numerous, more obtuse, and less prominently looped near the borders. Except for their more obtuse secondaries, the leaves of L. socialis are also comparable to those of L. utahensis Lesquereux ^ from the Raton formation. It seems probabie that at least the Denver and Medicine Bow speci- mens should be described as a new species. In the absence of better and more nuraerous speci- mens, however, a revision and description cannot be made at the present time. In addition to Evanston, Raton, Denver, and Dawson occurrences reported above, Laurus socialis is found also in the Livingston formation of Montana.^ Among modern leaves tlie lauraccous gcnera appear to resemble those of Laurus socialis most closely. Several species of Persea and Laxirus with similar leaf characters have been noted. Laurus nobilis Linn6, for example, has leaves of essentially similar sliape and venation. An even closer comparison may be noted in the leaves of Persea carolinensis (L.) Nees.' A general re- semblance is also apparent with leaves of Magnolia (Magnoliaceffi) and Diospyros (Ebenaceae). Occurrence — Elk Mountain road, Wyoming, Loc. P. 374. Collection—V. C. Mus. Pal., Plesiotype? No. 1335. Family TROCHODENDRACE.E Genus TROCHODENDROIDES Berry Trochodendroides nebrascensis (Newberry) Dorf, n. comb. (Plate 11, Figs. 1, 4, 6, 7) Populus nebrascensis Newberry, New York Lyceum Nat. Hist. Annals, vol. 9, 62, 1868; Illustrations of Creta- ceous and Tertiary Plants, pl. 12, figs. 4, 5, 1878; U. S. Geol. Surv. Mon. 35, 47, pl. 27, figs. 4, 5, 1898; Lesquereux, Harvard Coll. Mus. Comp. Zoology BuII., vol. 16, 47, 1888. Populus arclica Heer. Lesquereux, Harvard CoII. Mus. Comp. Zoology BuII., vol. 16, 48, 1888. Populus crenata Unger. Lesquereux, ibid., 48. Populus nebrascensis var. grandidenlata Lesquereux, ibid., 47; Ivnowlton, U. S. Geol. Surv. Prof. Paper 155, 56, pl. 19, figs. 4-7, pl. 20, figs. 2, 6, 7, 1930. Populus 7iebrascensis var. rolundala Lesquereux, op. cit., 47; Knowlton, op. cit., 56, pl. 20, figs. 1, 3-5, 8. Populus nebrascensis var. acute-dentata Lesquereux, op. cit., 47; Knowlton, op. cit., 56, pl. 21, figs. 1, 2. Populus nebrascensis var. longifolia Lesquereux, op. cit., 48; Knowlton, op. cit., 56, pl. 21, figs. 3-5. Populus tenuinervata Lesquereux, op. cit., 48; Knowlton, op. cit., 58, pl. 22, figs. 1-4, 6. Populus zaddachi Heer. Lesquereux, op. cit., 47; KJiowlton, op. cit., 60, pl. 22, fig. 8. Populus jacksoni Knowlton, op. cit., 58, pl. 21, figs. 6-8, pl. 22, fig. 7. Populus zeilkri (Lesquereux) Knowlton, ibid., 54, pl. 18, fig. 8, pl. 19, figs. 1-3. Proloficus zeilleri Lesquereux, Harvard Coll. Mus. Comp. Zoology Bull., vol. 16, 50, 1888. Cissus? cannoni Knowlton, U. S. Geol. Surv. Prof. Paper 155, 113, pl. 47, fig. 5, 1930. Cissus lesquereuxi Knowlton, ibid., 114, pl. 48, figs. 3, 4. The imposing synonjTiiy of this species is the continuation of the attempt to rcvise the status of the variable leaves usually referred to species of Populus. That this task is not completed is indicated in part by a recent list given to me by Dr. R. W. Brown of the U. S. Geological Survey, in which he tentatively assigns eleven additional species of the Denver flora to the above species. I have not included these species for want of complete specimcns of them for comparisons. ' Knowlton, F. H., tj. S. Geol. Surv. Prof. Paper 101, 318, pl. 90, fig. 4, 1917. 2 Knowlton, F. H., U. S. Geol. Surv. BuU. 105, 56, 1893. ' See sheet No. 481, Biscayne Bay, Florida, at the Princeton University herbarium. 62 CONTRIBUTIONS TO PAL^ONTOLOGY In the Medicine Bow coUections there are over 30 fairly complete specimens and several fragments which could have been either segregated into about a dozen species or combined into a single species of somewhat variable leaf form. In view of the intergrading characters of these specimens and their occurrence in great numbers at one horizon of Locahty P. 374, I beUeve it justifiable to refer them to one species. In the references given above, both Lesquereux and Knowlton have admitted that many of the species and varieties hsted were chfficult, if not impossi- ble, to differentiate from the original specimens, referred at that time to Populus nebrascensis. The reference of these species to the genus Trochodendroides rather than Populus is based upon the greater resemblance of the leaves to those of several trochodendraceous genera and the occurrence elsewhere of typical seeds and pods of that family in close association with the leaves of this type. Both Berry and Brown liave previously presented adequate evidence for the reference to Trochodendroides} The fossil specimens from the Medicine Bow formation resemble closely the leaves of the hving Tetracentron chinensis Ohver ^ of eastern Asia. The stratigraphic and geographic chstribution of the various species of Trochodcndroides, as now conceived, in the Cretaceous and Tertiary deposits of the Rocky Mountain region is still inadequately known. Trochodendroides nebrascensis appears to be most abundant in the lower Denver and lower Medicine Bow formations. Its reported occurrence in the Paskapoo (Fort Union?) of British Cokunbia ' cannot be corroborated because of the lack of figured specimens. Moreover, the Fort Union age of the beds from which the original type specimens were coUected is difhcult to substantiate because of the inadequate description of tlie locahty, which is reported by Newberry ^ as "Banks of Yellowstone River." That the species occurs in the Lance formation is indicated by U. S. National Museum specimens from the "Converse County beds" whicli Knowl- ton tentatively referred to Populus sp.^ My own collection from these same beds east of Lance Creek, Wyoming, contains several leaves which are definitely referable to Trochodendroides nebra- scensis. The species also is present in Ward's coUection of leaves from Black Buttes, Wyoming, which I have consulted at the U. S. National Museum.' Though somewhat variable in habit, the leaves liere referred to Trochodendroides nebrascensis are consistent in their ovate shape, rounded or obtuse base, toothed margin, and shghtly narrowed and elongate tip. In these respects they can readily be distinguished from the widespread Fort Union species, Trochodendroides cuneata (Newberry) Berry,' and from others, sucli as Populus anoniala Ward, P. cordata Newberry, P. craspedodronia Ward, P. glandulifera Heer, P. nervosa Newberry, P. newberryi Cockerell, P. richardsoni Heer, and P. speciosa Ward, all of which are reported from the Fort Union formation. The closest resemblance among Fort Union species is with P. daphnogeyioides Ward, which is, however, more narrowly ovate than the average leaf form of Trochodendroides nebrascensis. Pending the final revision of the Fort Union flora by Dr. R. W. Brown of the U. S. Geological Survey, no further attempt can be made to clarify the taxonomic allocation of the numerous leaves of this type. Occurrence — Corson Ranch, Wyoming, Loc. P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374; Craig, Colorado, Locs. P. 375, P. 376. Collection—U. C. Mus. Pal., Plesiotypes Nos. 1336, 1337, 1338, 1339. FamiJy MAGNOLIACE^ Genus MAGNOLIA Linn^ Magnolia lakesii Knowlton (Plate 10, Fig. 2) Magnolia lakesii Kiowlton, U. S. Geol. Surv. Prof. Paper 130, 142, pl. 13, fig. 2, 1922; Brown, U. S. Geol. Surv. Prof. Paper, manuscript, 1937. There are 8 well-preserved specimens in the collections which are referable to this species, which was originally described from the Laramie formation. As indicated above, Brown has ' Berry, E. W., U. S. Geol. Surv. Prof. Paper 129, 166, 1922; Canada Geol. Surv. Mem. 182, 34, 1935; Brown, R. W.. U. S. Geol. Surv. Prof. Paper, manuscript, 1937; Geol. Soc. Am., Prel. Abstr., 68, Deo. 1937. ' See sheet No. 2156 at The New York Botanieal Garden. •Penhallow, D. P., Rept. Tert. Pl. Brit. Col., 78, 1908. < Newberry, J. S., New York Lyceum Nat. Hist. Annals, vol. 9, 62, 1868. ' Knowlton, F. H., unpublished manuscript, U. S. National Museum. ' See specimen of Populus sp., No. 1149. ' Berry, E. W., Canada Geol. Surv. Mem. 182, 34, 1935. FOX HILLS AND LOWER MEDICINE BOW 63 recently found this spccies in the Colgate formation of eastern Montana. I havc con:iparcd the Medicine Bow specimens with botii tiie original typc specimens and the Colgate specimcns and find them all alike in essential characteristics. The reference to the genus Magnolia may be correct, though as yet unsubstantiated by asso- ciated seeds, flowers, or wood. I have seen no modern leaves closely resembUug these specimens. Occurrence — Elk Mountain road, Wyoming, Loc. P. 374. Collection—U . C. Mus. Pal., Plesiotype No. 1340. Magnolia dakotana Berry (Plate 10, Fig. 3) Magnolia dakotana Berry, U. S. Geol. Surv. Prof. Paper 185-F, 130, pl. 25, figs. 6, 7, 1934. There are 6 specimens of this type, of which the best is figured. Except for the slight ab- normahty in the basal venation of the figured type, it is difficult to cUfferentiate between the Lance and Medicine Bow specimens. Despite the well-defined characters of this species, there appear to be no other fossil forms with which a comparison can be made. Berry's reference to the genus Magnolia was made with some hesitation. At present it seems desirable to retain it there pending further studies. Occurrence — North of Walcott, Wyoming, Loc. P. 373; TCorson Ranch, Wyoming, Loc. P. 371. Collection^V. C. Mus. Pal., Plesiotype No. 1341. Magnolia nervosa (Knowlton) Brown (Plate 10, Fig. 4) Mag-nolia ncrvosa (ICnowlton) Brown, U. S. Genl. Surv. Prof. Paper, manuscript, 1937. Pterospermites nervosns Ivnowlton, U. S. Geol. Surv. Prof. Paper 101, 273, pl. 48, fig. 4, 1917. This type of leaf is rare in the Medicine Bow coUections, being represented only by a few fragments and the single fairly complete leaf figured. This specimen is identical with those from the Colgate and lower Lance formations of eastern Montana which are in the collections of the U. S. National Museum. Brown has referred these to Magnolia yiervosa, with which he has syn- onymized the specimen of Pterospermites nervosus from the Vermejo formation. Occurrence — Elk Mountain road, Wyoming, Loc. P. 374. Collection—\] . C. Mus. Pal., Plesiotype No. 1342. Magnolia pulchra Ward (Plate 11, Fig. 8) Magnolia pulchra Ward, U. S. Geol. Surv., 6th Ann. Rept., 556, pl. 60, figs. 2, 3, 1886; idem, Bull. 37, 103, pl. 48, figs. 3, 4, 1887. This species is represented only by several fragmentary leaves and the nearly complete specimen figured. It is identical in all respects with the type and figured specimen of Magnolia pulchra, except possibly for the tip, which is turned under in the Medicine Bow specimen but ap- pears to be slightly more obtuse. The obovate shape and thin secondaries clearly distinguish this species from the others referred to the same genus. The original specimens of this species are from the' Mesaverde formation. Berry has recently reported it from the upper Ravenscrag (Fort Union) of southern Saskatchewan.^ I have not been able to see his specimens for comparison, so cannot comment on their relationship to the late Cretaceous specimens. The generic reference to Magnolia seems reasonable. I have noted also a resemblance of the specimens to the leaves of Asimina triloha Dunal (Anonaceje), which are identical in shape and size and differ onJy slightly in their more regular secondary veins. Occurrence — Corson Ranch, Wyoming, Loc. P. 371; Elk Mountain road, Wyoming, Loc. P. 374. Collection—V. C. Mus. Pal., Plesiotype No. 1343. 1 Berry, E. W., Canada Geol. Surv. Mem. 1S2, 35, 1935. 64 CONTRIBUTIONS TO PAL^ONTOLOGY Genus MAGNOLIOPHYLLUM Conwentz MagnoliophyUum cordatum Dorf, n. sp. (Plate 9, Fig. 5; Plate 10, Fig. 1) This type of leaf is fairly abundant, although its large size has made it impossible to procure complete specimens. The figured specimen is the most nearly complete of over 80 specimens collected in the field. Description. — Leaves cordate to broadly ovate, symmetrical, with the base cordate to trun- cate and the tip acute; length ranging from 10 to 18 cm., widtli from 8 to 16 cm., petiole stout and up to 2.5 cm. long; midrib very stout; secondaries well defined, ten to twelve pairs mostly alternate, leaving the midrib at angles varjdng from slightly more than 90° near the base, to 60° at the middle, and to 50° or less near the tip; basal secondaries only slightly arched upward, upper secondaries arching upward more abruptly, producing a greater distance between the secondaries near the margins than at the midrib; one to three prominent tertiaries branching from lower sides of secondaries forming conspicuous loops with superjacent tertiaries or secondaries close to the margins; remaining tertiaries strong, percurrent, forming a coarse network in the intersecondary spaces; margin mainly entire, occasionally undulate in the upper half of large leaves; texture firm. It seems strange that this well-defined type of leaf has apparently not been encountered previously in the late Cretaceous deposits of North America. I believe, however, that it is closely related to the specimens from the Raton formation which Knowlton referred to Magnolia cordifolia Lesquereux.^ There are notable differences, however, chiefly in the character of the base, the upper secondaries, and the number and prominence of the marginal loops, all of which are con- sistently the same in Magnoliophyllum cordatum. I have not been able to determine the taxonomic affinities of this species, although its char- acters are strongly suggestive of the family Magnoliacese. The smaller leaves of the living Magnolia dealbata Zucc. of Mexico and the larger leaves of M. acuminata Linne of southeastern United States ^ are of the same size and have almost identical venation and basal characters. I have found no other modern leaves which are so nearly identical, so do not hesitate in referring the species to Magnoliophyllum, implying relationship to the family Magnohacese. Occurrence — Corson Ranch, Wyoming, Locs. P. 371, P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374; Craig, Colorado, Loc. P. 375. Collection—V. C. Mus. Pal., Cotypes Nos. 1344, 1345. FAMILY LEGUMINOSiE Genus LEGUMINOSITES Bowerbank Leguminosites arachioides minor Berry (Plate 11, Figs. 2, 3, 5) Leguminosiles arachioides minor Berry, U. S. Geol. Surv. Prof. Paper 156, 89, pl. 14, figs. 2-6, 1930; Canada Geol. Surv. Mem. 182, 64, 1935. Leguminosites? arachioides (Lesquereux) Lesquereux. Penhallow, Rept. Tert. Pl. Brit. Col., 61, fig. 14, 1908. Nyssa (?) racemosa Kjiowlton, U. S. Geol. Surv. Bull. 152, 153, 1898; idem, Prof. Paper 134, 95, 1924. Berrya racemosa Knowlton, U. S. Geol. Surv. Prof. Paper 155, 134, pl. 41, figs. 4-5, 1930. Sabalites fructifer Lesquereux (in part), Rept. U. S. Geol. Surv. Terr., vol. 7, 114, pl. 11, figs. 3 (fruits at base of figure only), 3a, 1878. This type of seed is represented in the collections by a compound raceme and its counterpart and several detached fruits. Berry described specimens from the Wilcox formation as the variety minor of the well-known and widespread Legmninositesf arachioides (Lesquereux) Lesquereux, on the basis of their smaller size. The Medicine Bow specimens appear to be slightly smaller still, though of the same size exactly as several detached fruits from the Wilcox and Denver. In every other respect our speci- mens are identical with the type and figured specimens from the Wilcox, the Raton, and the Denver formations. The species is also recorded from the Animas, Dawson, and Black Buttes floras. ' Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 101, 315, pl. 86, pl. 89, fig. 1, 1917. ' See sheet No. 13, Princeton University herbarium. FOX HILLS AND LOWER MEDICINE BOW 65 The systpmatic statiis of tliis typo of spchI is stiil unsettlod. Lesqucreux regarded it as a palm, and Knowlton was ineliued toward Nysm, thougii he later reconsidered and gave it the non- committal name Berrya, apparently unaware that this name was preoccupied and that Berry was at that precise time describing the same type of seed from the Wilcox beds as Leguminosites arachioides minor, implying leguminous afhnities. Quite recently Brown ' lias pointed out the trochodendraceous character of the fossil specimens and their visual association, as in the Medicine Bow formation, witli leavcs referred to Trochodendroides. Occurrence — Corson Ranch, Wyoming, Loc. P. 372; north of Walcott, Wyoming, Loc. P. 373. Collection—V . C. Mus. Pal., Plesiotypes Nos. 1346, 1347, 1348. FAMILY ANACARDIACE^ Genus PISTACIA Linn6 Pistacia eriensis Knowlton (Plate 19, Fig. 7) Pistacia eriensis Knowlton, U. S. Geol. Surv. Prof. Paper 130, 150, pl. 28, figs. 1-4, 1922. Pistacia hollicki Knowlton, ibid., 151, pl. 28, figs. 5, 6. There are 4 leaves with attached leaflets and 3 detached leaflets whicli are indistinguishable from the Laramie specimens referred to Pistacia eriensis. Two of the spccimens, one of which is figured, are somewhat larger than any of the Laramie specimens but are otherwise similar. In view of the very slight differences in size and arrangement between Pistacia eriensis and P. hollicki, and the fact that the speciraens all come from one locality, there seem to be no grounds for maintaining two distinct species. Knowlton referred to them as "obviously related and possibly identical." ^ No other fossil forms appear to be comparable, although the iarge leaves from the same Laramie locality which Knowlton described as Juglans newberryi ' might ultimately prove to be variants of Pistacia eriensis, since their shape and venation are remarkably .similar. Knowlton referred this species to the genus Pistacia with considerable certainty, on the ba.sis of similar arrangement, shape, and venation to Pistacia lentiscns Linne and P. mutica Frick and Meyer of Southern Europe. The re.semblance is indeed close, although equally good comparisons can be made with species of Rhus, belonging to the same family, and with Juglans of the Jug- landacese. In the absence of conclusive evidence, however, it does not seem desirable to make any change in the name at the present time. Occurrence — Corson Ranch, Wyoming, Locs. P. 371, P. 372; north of Walcott, Wyoming, Loc. P. 373. Collection—U . C. Mus. Pal., Plesiotype No. 1349. Family CELASTRACE^ Genus CELASTRUS Linn6 Celastrus taurinensis Ward (Plate 12, Figs. 1, 2, 3) Celaslrus laurinensis Ward, U. S. Geol. Surv., 6th Ann. Rept., 555, pl. 52, figs. 15, 16, 1886; idem, Bull. 37, 79, pl. 34, figs. 5, 6, 1887; Hollick, Geol. Surv. Louisiana, Spec. Rept. No. 5, 285, pl. 46, fig. 1, 1899; Knowlton, Wasliington Acad. Sci. Proc., vol. 11, 213, 1909; Berry, U. S. Geol. Surv. Prof. Paper 91, 267, pl. 60, figs. 1-3, 1916; Canada Geol. Surv. Meni. 182, 42, 1935. Celastrus ovalus Ward, U. S. Geol. Surv., 6th Ann. Rept., 555, pl. 53, fig. 7, 1886; idem, Bull. 37, 81, pl. 36, fig. 1, 1887. Celaslrus wardii Knowlton and Cockerell, U. S. Geol. Surv. Bull. 696, 160, 1919. Celaslrus curvinervis Ward, U. S. Geol. Surv., 6th Ann. Rept., 555, pl. 53, figs. 9, 10, 1886; idem, BuU. 37, 82, pl. 36, figs. 3, 4, 1887. Aralia taurinensis (Ward) Sanborn, Carnegie Inst. Wash. Pub. No. 465, 27, pl. 10, figs. 1, 2, 4, 1935. The coUections contain 5 well-preserved and fairly complete specimens which are similar in every essential detail to the type and figured specimens of Celastrus wardii (= Celastrus ovatus) and C. curvinervis. Berry has recently ^ synonymized both of these species with Celastrus tauri- 1 Brown, R. W., Geol. Soc. Am., Prel. Abstr.. 68, Dec. 1937. ' Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 130, 151, 1922. 'Knowlton. F. H.. ibid., 120. pl. 20, figs. 8-10. * Berry, E. W., Canada Geol. Surv. Mem. 182, 42, 1935. 66 CONTRIBUTIONS TO PAL.EONTOLOGY nensis, which Sanborn, at about the same date,' considered more appropriately referred to the genus Aralia. It seems to me that the species might equally well be a Viburnum. In my private herbarium I have a number of leaves of the Asiatic Viburnum seiboldiif, whose characters, particu- larly of venation and marginal teeth, appear to be more Hke those of the fossil specimens than are those of any Aralia. In the range of size and shape variations, moreover, these Viburnuin leaves can exactly match the variations seen in the leaves of Celastrus taurinensis and its synonynious forms. In view of the scarcity of leaves of this type in the Medicine Bow collections, however, it does not seem desirable to make further changes in the taxonoray, but rather await the revision of the Fort Union flora by Dr. Roland W. Brown, who has a large suite of specimens of this species at his disposal at the U. S. National Museura. This species is evidently widespread and of long geologic range in North America. In addition to its recorded occurrences, there are also several perfect specimens of it in the unclassified material from Black Buttes, Wyoming, which I have consulted at the U. S. National Museum. It is one of the few Medicine Bow species which ranges up into the Fort Union of the Rocky Mountain region. As seen in the citations above, it is also common in the Eocene of the Gulf Coast and of Oregon. Any attempt to separate the late Cretaceous forras from those of the Paleocene and Eocene on justifiable grounds has been futile, although the Cretaceous specimens do seem to be more lanceolate in average shape. Whether or not this character can become the basis for specific separation remains to be seen in subsequent studies. Occurrence — Corson Ranch, Wyoming, Locs. P. 371, P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374. Collection—\5 . C. Mus. Pal., Plesiotypes Nos. 1350, 1351, 1352. Family RHAMNACE^ Genus RHAMNUS Lmn6 Rhamnus salicifolius Lesquereux (Plate 12, Figs. 4, 5; Plate 13, Fig. 4) Rhamnus salicifoliiis Lesquereux, Am. Jour. Sci., 2d ser., vol. 45, 206, 1868; U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1869), 196 (reprint, 1873); Rept. U. S. Geol. Surv. Terr., vol. 7, 282, pl. 53, figs. 9, 10, 1878; Ivnowlton, U. S. Geol. Surv. Prof. Paper 101, 271, 1917; idem, Prof. Paper 130, 154, pl. 15, fig. 4, pl. 19, fig. 2b, 1922. Rhamnus salicifolius Lesquereux?. Ivnowlton, U. S. Geol. Surv. Bull. 163, 70, 1900; idem, Prof. Paper 155, 107, pl. 46, fig. 12, 1930. Rhamnus bebrwntensis Knowlton and Cockerell, U. S. Geol. Surv. Bull. 696, 544, 1919; ICnowlton, U. S. Geol. Surv. Prof. Paper 130, 155, 1922. Rhamnus elegans Newberry, New York Lyc. Nat. Hist. Annals, vol. 9, 49, 1868; U. S. Geol. Surv. Mon. 35, 117, pl.'50, fig. 2, 1898. Juglans larandensis Knowlton, U. S. Geol. Surv. Prof. Paper 130, 120, pl. 20, fig. 12, 1922. Ficus? smiihaoniana? (Lesquereux) Lesquereux. Knowlton, ibid., 130, pl. 21, fig. 4. Lanceolate leaves, referable to Rhamnus salicifolius, are abundant and widespread in the Medicine Bow formation. More than 20 well-preserved specimens in the collections indicate considerable variation in size, yet are identical in secondary and tertiary venation and general shape. The average characteristics are precisely as shown in Lesquereux's original figures,^ except that the reconstruction of the broken tip of his figure 9 should apparently have been con- siderably elongated. In view of the intergrading variations observed in the Medicine Bow speci- mens, it is evident that the species listed above as synonymous are merely shght variants of the average leaf forra of Rhamnus salicifolius. It seems desirable to re-describe this species on the basis of the abundant material collected. Description (Supplementary) — Leaves typically narrowly lanceolate, narrowing gradually from below the middle to an elongate acute tip, and raore abruptly below to an equilateral or shghtly inequilateral cuneate base; length 5 to 12 cm., width 1.5 to 2.5 cm., average dimensions apparently about 10 by 2 cm., petiole short, thin, up to 1 cm. long; midrib firm, relatively straight; 10 to 14 pairs of thin, mainly alternate secondaries leaving the midrib at angles of 25° to 45°, curv- ing uniformly upward toward the margins, converging, unbranched, paralleling margins and > Sanborn, Ethel I.. Carnegie Inst. Wash. Pub. No. 465, 27, 1935. 'Lesquereux, Leo, Rept. U. S. Geol. Surv. Terr., vol. 7, pl. 53, figa. 9, 10, 1878. FOX HILLS AND LOWER MEDICINE BOW 67 beeoming indistinet close to the leaf borders; tertiaries thin, percurrent at angles approachinp; 90° to tlie plane of the midrib; finer nervation not prcserved; margin entire throuKhout; texture firm. This species appears to be most abundant in the Laramie and Medicine Bow formations. It is evidently lcss abundant though definitely present in thc Vermejo and Black Buttes fforas, and is questionabiy idcntified in the Denver and Dawson floras. Its reported occurrence in the Lance formation ' seems likeiy, thouRh as yet unsubstantiatcd by fisured specimens. To my knowledge tliere are no reported occurrences of leaves of this type from formations of indisputable Paleocene or later age. The leaves of this fossil spccies are, in my opinion, distinctly not rhamnaceous. I have studied all available hcrbarium material of the genera of the Rhamnaceie at The New York Botanical Garden and failed to find a comparable leaf type. Scveral genera of the Lauracctc, on the other hand, are very simiiar in most details. Persea lanceolata has leaves of the same general shape and venation. Even closer are the leaves of Nectandra pichurim (H.B.K.) Mez.,' of northern South America, and A''. memhranacea (S.W.) Griesb.,' of Central America, which possess the same well-defined secondary and tertiary venation and the same variations in size and shape as the fossil specimens. Occurrence — Cor.son Rancli, Wyoming, Locs. P. 371, P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374; Craig, Colorado, Loc. P. 376. Collection—V . C. Mus. Pal, Plesiotypes Nos. 1353, 1354, 1355. Rhamnus cleburni Lesquereu.\ (Plate 14, Figs. 5, 6) Rhamnus clehurni Lesquereux, U. S. Geol. Surv. Terr., 6th Ann. Rept. (1872), 381, 400, 1873; Rept. U. S. Geol. Surv. Terr., vol. 7, 280, pl. 53, figs. 1-3, 1878; Berry, U. S. Geol. Surv. Prof. Paper 91, 283, 1916; Knowlton, U. S. Geol. Surv. Prof. Paper 101, 332, pl. 113, fig. 3, 1917; idem, Prof. Paper 15.5, 104, only pl. 46, figs. 10, 11, 1930. Rhamnus reciinervis Heer (in part). Lesquereux, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1871), 29.5, 298; idem (1872), 382, 397, 402, 1873; idem (1873), 405, 1874; Rept. U. S. Geol. Surv. Terr., vol. 7, 278, pl. 52, fig. 15 only, 1878. Rhamnus hrilloni Knovvlton, U. S. Geol. Surv. Prof. Paper 130, 156, pl. 15, fig. 6, pl. 24, fig. 8, 1922. This type of leaf is one of the more abundant and widespread in the Medicine Bow collections. It is remarkably consistent in the majority of its observable characters, being particularly so in its secondary and tertiary venation. The average leaf specimen is identical with the original type specimens of Rhamnus cleburni from the Denver formation. In studying the large number of leaves of this type in my collection, however, it has become apparent that the leaves are somewhat more variable, as one would expect, than Lesquereux's description and figures might imply. The two specimens here figured were chosen to show such variations rather than the average type. In figure 5, for example, is shown a somewhat more lanceolate leaf than the average, whose upper margin is invested with short, blunt, upward-pointing teeth, identical with those seen in Les- quereux's figure of Rhamnus rectinervis. The largest specimen obtained is shown in figure 6. This is somewhat wider above the middle than the average form, and has a more elongate tip; otherwise it is precisely the same. In view of the range of variations now known it seems desirable to re- describe the species as here conceived. Description {Supplementary) — Leaves oval or elliptic, more rarely lanceolate, the shghtly unequal base narrowly to broadly cuneate, rarely rounded, and the apex acute or slightly extended into an acuminate tip; average dimensions about 11 cm. by 4.5 cm., ranging up to 17 cm. by 9 cm.; minimum size observed in collections 7 cm. by 4 cm.; petiole stout, relatively long, up to 3 cm.; midrib straight, heavy except near the apex where it becomes thin and occasionally undulating; 9 to 15 pairs of prominent secondaries, subopposite to alternate, leaving the midrib at angles consistently from 40° to 45° except near the apex where they become more acute; rarely the basal pair are slightly less acute; secondaries closely and evenly set along the niidrib, becoming more 1 HoUick, A., in Brown, B., Am. Mus. Nat. Hist. Bull., vol. 23, 823-845, 1907; Knowlton, F. H., Washington Acad. Sci. Proc. vol. 11, 18.5, 1909. » See sheets No. 1396 and No. 4323, New York Botanioal Garden. ' See sheets No. 987 and No. 169, New York Eotanical Garden. 68 CONTRIBUTIONS TO PALiEONTOLOGY remote near the apex, consistently parallel, curving only slightly toward the margins where they curve upward more abruptly and run into the margins; secondaries near apex parallehng and curving in toward the midrib; tertiaries fine and very closely spaced, flexuose percurrent, at ap- proximately right angles to the secondaries; very rarely, tertiary branches given off from the lower sides of the secondaries near the margins; finer nervation not observable; margin usually entire, slightly undulate, more rarely with blunt, inconspicuous, upward-pointing teeth near the apex; texture firm. The characters of this type of leaf are so easily recognized and consistent that I feel no hesi- tancy in referring Rhamnus brittoni and one specimen of R. rectinervis to the same species. I was at first also incUned to include in it the specimens referred to R. goldiajius,^ but find that in none of my specimens is there any trace of the well-developed branching of the lower secondaries found in that species. Knowlton has remarked upon the similarity of certain specimens from tlie Denver flora referred to Berchemia multinervis (Al. Braun) Heer.^ These are, however, of chfferent shape and do not possess the characteristic upper secondaries parallchng and curving in toward the midrib. In the same report, Knowlton included in Rhavmus cleburni the specimens which Lesquereux had referred to Cornus studeri Heer?. These differ, however, in their more remote, converging secondaries, which are usually forked near the leaf margins. As previously recorded, Rhamnus cleburni and the species here included in it are known from the Laramie, Denver, Dawson, Raton, and BLack Buttes floras of the Rocky Mountain region. It is also present in coUections from the Lance of " Converse County," Wyoming, both at Princeton University and the U. S. National Museum. It is not known to occur in the Fort Union floras of western North America, but is reported, though unfortunately not figured, from the Eocene of Oregon' and the Gulf Coast.'' Berry has pointed out "the similarity and possible identity" of the Gulf Coast specimens to the variable leaves of Rhamnus coushatta Berry.'* It is in any event apparent that both the Gulf and West Coast Eocene species are closely related to the late Cre- taceous R. cleburni, indicating a survival of descendant forms in regions outside the Rocky Moun- tain area. It is difficult to see any resemblance between the fossil leaves of Rhamnus clehurni and the leaves of modern species of Rhamnus. In my opinion, after consulting herbarium specimens at The New York Botanical Garden, the fossil specimens are not referable to any hving genus of the family Rhamnacese. On the other hand, their leaf characters are closely similar to several genera of the family Dilleniaceaj, namely Davilla, Dillenia, Doliocarpus, and Tetracera. In Dillenia and Tetracera the leaves are usuaUy more regularly toothed along the whole margin than in the fossil specimens. The leaves of Doliocarpus de^Uatus (Aubl.) Standley " of Central and South America are practicaUy indistinguishable in aU details from the fossil specimens. Almost equaUy com- parable are the leaves of Davilla multiflora St. Hil.,' which differ only in their shghtly more obtuse secondaries. Other genera which were consulted but eliminated because of marked differences were Rhamnidium (Rhamnaceaj), Cuphea (Lythraceae) and Meliosma (Sabiaceee). Occwrrence— Corson Ranch, Wyoming, Loc. P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374; Craig, Colorado, Loc. P. 376. Collection^V. C. Mus. Pal., Plesiotypes Nos. 1356, 1357. Genus ZIZYPHUS Adanson Zizyphus hendersoni Knowlton (Plate 12, Fig. 6) Zizyphus hendeTSoni Knowlton, U. S. Geol. Surv. Prof. Paper 130, 157, pl. 15, figs. 1, 2, 1922. Zizyphus coloradensis Knowlton, ibid., 157, pl. 15, fig. 5. There are 5 specimens in the coUections which were at first thought rcfcrable to Ficus planicostata, but which are consistently different in their thinner texture and the presence of 5 > Lesquereux, Leo, Rept. U. S. Geol. Surv. Terr., vol. 7, 281, pl. 53, figs. 4-8, 1878; Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 101, 332, pl. 101, fig. 4, pl. 112, fig. 5, 1917. 2 Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 155. 104, 1930. > Knowlton, F. H., U. S. Geol. Surv. Bull. 204, 80, 1902. * Berry, E. W., U. S. Geol. Surv. Prof. Paper 91, 283, 1916. ' Berry, E. W., U. S. Geol. Surv. Prof. Paper 131, 15, 1923. ' See eheet No. 3748 at The New York Botanical Garden. ' See sheet No. 5359. FOX HILLS AND LOWER MEDICINE BOW 69 rather than 3 basal primarics. Thcy cannot be distinguished from the Laramie specimcns which Knowlton called Zizyphu.s hendersoni. The spccimcn of Z. coloradensin , which I have examined at the U. S. National Museum, is clcarly thc same species, since it does not show any trace of upper marginal teeth, as Knowlton suspected. This species is at present known only from the Laramie and Medicine Bow formations. 1 have been unable to find any modern leaves with which this species can be compared. Occurrcnpe — Corson Ranch, Wyoming, Loc. P. 372; north of Walcott, Wyoming, Loc. P. 373; Elic Mountain road, Wyoming, Loc. P. 374. Collection—\J . C. Mus. Pal., Plesiotype No. 1358. Family VITACE^ Genus CISSITES Heer Cissites lobatus Dorf, n. sp. (Plate 13, Fig.s. 2, 3) In the field collcctions there were 27 specimens of this type, which were consistent in shape, average size, and venation. Four of the better specimens were retained for study. Descri-ption — Leaves oval to elliptic, widest just below the middle, obtusely wedge-.shaped at the base, and gradually narrowed to a slightly elongate, acuminate tip; length 6 to 9 cm.; width 3 to 5 cm.; petiole at least 1.2 cm. long; secondaries tri-nerved from just below the base of the blade, the laterals almost as heavy as the midrib, and given off at an angle of 20°; next pair of secondaries branching from the midrib just below the middle; these and the widely spaced sec- ondaries above are acute, sUghtly curved upward, and forking toward the margin forming loops with the secondaries or their branches above; 3 to 4 prominent forking tertiaries along the lower edge of the lateral basal secondaries, and usually one tertiary well developed from the upper edges; margin entire, palmately lobed by 2 pairs of shallow, narrow, rounded sinuses; central lobe largest, symmetrically oval; lateral lobes smaller and asymmetrical ; texture subcoriaceous. This species shows only a general resemblance to other fossil species of Cissites from North America. The closest similarity is with C. colgatensis Brown, which, however, is uniformly smaller, more distinctly lobed, and of an entirely different shape.i The resemblance to C. panduratus Knowlton from the Vermejo formation is also apparent though more remote.^ It is, however, similar to specimens from the Lance of "Converse County," Wyoming, which Knowlton tenta- tively called Ficusf lobata Knowlton in an unpublished manuscript at the U. S. National Museum. In the attempt to determine the probable botanical affinities of this type of leaf I have sur- veyed all available herbarium material at The New York Botanical Garden of the following fami- lies: Arahaceffi, Magnohaceie, Moracese, Stercuhaceee, and Vitacete. The practice of referring leaves of this general form to Cissites, implying relationship to the genus Cissus (Vitacese), is in a measure substantiated by the leaf characters of several Uving species of that genus, notably C. rhombifolia (sheet No. 557). A general resemblance was also seen to the variable leaves of Pterospermum proteus BurkiU (sheet No. 10120) of the Sterculiacea;. No exact dupUcate of the fossil specimens has yet been seen in any of the families studied. Occurrence — Corson Ranch, Wyoming, Loc. P. 372. Collection—lJ. C. Mus. Pal., Type No. 1359, Paratype No. 1360. Genus VITIS (Touraefort) Linn6 Vitis stantoni (Ivnowlton) Brown (Plate 17, Fig. 1) Vitis slantoni (Knowlton) Brown, U. S. Geol. Surv. Prof . Paper, manuscript, 1937. Castalia stanloni Knowlton, U. S. Geol. Surv. Bull. 257, 147, pl. 19, fig. 4, 1905. Viburnum vulpinum Knowlton, U. S. Geol. Surv. Prof. Paper 98, 92, pl. 18, fig. 1, 1917. Vitis dakolana Berrj', U. S. Geol. Surv. Prof. Paper 185-F, 130, pl. 26, figs. 4-6, pl. 27, 1934. There are two specimens from Locality P. 374 which appear to be referable to this species, which Brown has recently sjmonymized as given above. As now known this species occurs in ' Brown. R. W., U. S. Geol. Surv. Prof. Paper, manuseript, 1937. 2 Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 101, 274, pl. 49, fig. 10, 1917. 70 CONTRIBUTIONS TO PAL^ONTOLOGY the Judith River (?), Fox Hills, Colgate, Lance, and Medicine Bow floras. It is apparently most abimdant in the lower Lance formation, from which I have collected a considerable number of perfect specimens near Lance Creek, Wyoming, and near Buffalo, South Dakota. The reference to the genus Vitis seems reasonable though as yet not corroborated by asso- ciated seeds, flowers, or wood. Occurrence — Elk Mountain road, Wyoming, Loc. P. 374. Collection—lJ. C. Mus. Pal., Plesiotype No. 1361. Family TILIACEiE Genus APEIBOPSIS Heer Apeibopsis? laramiensis Knowlton (Plate 14, Fig. 1) Apeibopsis? laramiensis Knowlton, U. S. Geol. Surv. Prof. Paper 130, 159, pl. 7, fig. 4, 1922. There are 2 well-preserved, incomplete specimens from Locality P. 374 which are similar except for their somewhat smaller size to the type and figured specimens of this species from the Laramie formation. Its very truncate base, thick midrib, and thin, closely spaced, widely diver- gent secondaries are easily recognizable and diagnostic. The generic reference has been justifiably questioned by ICnowIton. Until more complete specimens are found, however, it is difficult to attempt a taxonomic revision. Occurrence — Elk Mountain road, Wyoming, Loc. P. 374. Collection—\J. C. Mus. Pal., Plesiotype No. 1362. Genus GREWIOPSIS Saporta Grewiopsis saportana Lesquereux (Plate 13, Figs. 1, 5, 6) Grentiopsis saporlana Lesquereux, Rept. U. S. Geol. Surv. Terr., vol. 7, 257, pl. 50, figs. 10-12, 1878; Brown, U. S. Geol. Surv. Prof. Paper, manuscript, 1937. Aleurites eocenica Lesquereux, U. S. Geol. Surv. Terr., 6th Ann. Rept. (1872), 397, 1873. Greu-iopsis ficifolia Ward, U. S. Geol. Surv., 6th Ann. Rept., 556, pl. 46, figs. 1, 2, 1886; idem, Bull. 37, 92, pl. 41, figs. 1, 2, 1887. Grewiopsis eocenica (Lesquereux) Ivnowlton, U. S. Geol. Surv. Bull. 152, 114, 1898; Wasliington Aead. Sci. Proc, vol. 11, 207, 1909. This type of leaf is by far the most abundant at Locahty P. 374. Over 25 well-preserved specimens were collected for study. The majority of these cannot be distinguished in any essential detail from the type specimens from Black Buttes, Wyoming. These were originally described as Aleurites eocenica by Lesquereux, who subsequently changed the name to Greiviopsis saportana. In a later report Knowlton changed the name again to Grewiopsis eocenica, a procedure which hardly seems justified, as has been pointed out by Brown. Knowlton also included Ward's specimens of Grewiopsis ficifolia in this species, a reference clearly substantiated by the numerous Medicine Bow specimens. In view of the large number of variants now available it seems appro- priate to re-describe the species. Description {Supplementary) — Leaves elliptic-lanceolate to ovate, narrowing gradually to a blunt, obtuse point and more abruptly to a symmetrical, cuneate base; length 4.5 to 9.5 cm. ; width 2.2 to 6.5 cm., widest below the middle; average dimensions apparently about 7.5 by 4.5 cm.; petiole relatively thin, very long, up to 5 cm.; midrib heavy except near apex, fairly straight; 6 to 9 pairs of secondaries, mainly alternate, pinnate, approaching palmateness in basal pair, leaving midrib at angles of about 30° near apex to as high as 60° ncar base ; venation craspedodrome, the secondaries irregularly curved toward the margins where they usually branch into the mar- ginal teeth, these branches being more numerous in lower secondaries; tertiary venation distinct, uniform, the veins continuous or branched at angles approaching 90° to the secondaries; margin finely serrate or denticulate, often inconspicuously; the teeth short, irrogular in size, and usually closely spaced; texture subcoriaceous. FOX HILLS AND LOWER MEDICINE BOW 71 This species is apparently most abundant in the Black Buttos and Medicino Bow floras. Its reported occurrence in the Lanco formation of eastcrn Wyoming by Knowlton ' has been amply substantiated by his unpubhshed manuscript and figures at tho U. S. National Museum. In the summer of 1936, I visitcd this region and coilected several completc specimens clearly belonging to this same spccies. Brown has also coUected excellent specimens of it in the Hell Creek member of the Lance near Marmarth, North Dakota. There do not appear to be any other reported occurrences of the spocies in North America, although Chaney rcports that Callichlamys zeteki from the upper Eocene (?) of Oregon - is of the same general type. Except for their resemblance to leaves of the Uving genus Callichlamys, as pointed out by Chaney, I have found no other forms except Grewiopsis witli which the fossil leavos can be compared. Occurrence — Corson Ranch, Wyoming, Loc. P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374. Collection—V. C. Mus. Pal., Plosiotypes Nos. 1363, 1364, 1365. Family CORNACEiE Genus CORNOPHYLLUM Newberry Cornophyllum wardii Dorf, n. sp. (Plate 16, Fig. 3) Comus sluderi? Heer. Lesquereu.x, Rept. U. S. Geol. Surv. Terr., vol. 7, 244, pl. 42, figs. 4, 5, 1878; Ward, U. S. Geol. Surv., 6th Ann. Rept., 553, pl. 48, fig. 1, 1886; idem, Bull. 37, 55, pl. 26, fig. 1, 1887; Knowlton, U. S. Geol. .Surv. Bull. 163, 68, pl. 15, fig. 3, 1900; idem, Prof. Paper 101, 342, pl. 109, fig. 2, 1917; idem, Prof. Paper 155, 104, only pl. 40, fig. 6, 1930. The single, nearly complete specimen figured is one of 6 well-preserved specimens of this type of leaf. It is impossible to distinguish from the Mesaverde specimen referred by W^ard to the European species Cornus studeri. Knowlton has reviewed in detail the status of the various specimens wliich have been referred to this species, referring several to Rhamnus cleburni Les- quereux.' He adds, however, that Ward's Mesaverde specimen and his own specimens from the same formation are not referable to Rhamnus cleburni nor are they similar to the European Cornus studeri. W\t\\ the additional matorial now at hand it seems advisable to describe the species as new, despite the indefinite generic reference. Description — Leaves narrowly ovate to elhptic, gradually narrowed to an extended, slightly acuminate or acute tip, and more abruptly bolow to a widely cuneate base; longth 7 to 13 cm., width 4 to 7 cm., widest below the middle, midrib very thick bclow, thinning gradually to the apex; petiole thick, up to 1.5 cm. long, secondary venation pinnate, mainly alternate, camptodrome; secondarios more widoly spaced along midrib and more abruptly curved upward in uppor half than in lowor half of leaf, outer onds of secondaries tending to parallel margins for considerable distance, looping indistinctly with tcrtiary branches of secondaries next above; tortiary venation indistinct, nervilles running mainly at high angle to midrib; margin entire; texturo coriaceous. This species is of the general type of Rhamnus cleburni, though clearly distinguishable by the more widely spaced, less numerous secondaries, which do not evon approach the type of parallel venation of R. cleburni. There are also discernible cUfTerences in tertiary venation and general shape. It seems ovidont, therefore, that Lesquereux's Donver specimens of Cornus studeri? which Knowlton identified as Rhamnus cleburni bolong rather to the type here called Cornophyllum wardii. Also, one of Knowlton's spocimens of Rhamnus cleburni from the Denver flora (plate 40, figure 6) is clearly referable to Coriiophyllum ivardii. In addition to the Aledicino Bow occurrences, and its presence in the Mesaverde, Raton, and Denver formations, this species is also present in my colloctions of Lance material obtained in 1936 from two locahties east and northeast of Lance Croek, Wyoming. The general shape, entire margin, and secondary venation of this species resemble the leaves of the Uving species of Cornus, to which the previously recorded specimens wore roferred by ' Knowlton. F. H., Washington .\cad. Sci. Proc, vol. 11, 207, 1909. = Chaney, R. W., Carnegie Inst. Wash. Pub. No. 439, 95, 1933. ' Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 155, 105, 1930. 72 CONTRIBUTIONS TO PAL^ONTOLOGY Lesquereux, Ward, and Knowlton. I am foUowing the practice of Newberry and Berry in applying the less positive name Cornophyllum to the specimens. In addition to a resemblance to leaves of Cornus, the fossil leaves are also comparable in general form and venation to Rhamnidium (Rham- nacese), Berchemia (Rhamnaceae) and Premna (Verbenacea). Occurrence — Elk Mountain road, Wyoming, Loc. P. 374; Craig, Colorado, Loc. P. 376. Collection—lJ. C. Mus. Pal., Type No. 1366. Family BIGNONIACE^ Genus DOMBEYOPSIS Unger Dombeyopsis obtusa Lesquereux (Plate 15, Figs. 1, 2) Dojnbeyopsis obtusa Lesquereux, U. S. Geol. Surv. Terr., 6th Ann. Rept. (1872), 375, 1873; Rept. U. S. Geol. Surv. Terr., vol. 7, 255, pl. 47, figs. 4, 5, 1878; Knowlton, U. S. Geol. Surv. Prof. Paper 130, 162, pl. 13, fig. 4, pl. 20, fig. 11, pl. 27, figs. 1-4, 1922; idem, Prof. Paper 155, 124, 1930. Phylliles populoides Knowlton, U. S. Geol. Surv. Prof. Paper 101, 280, pl. 50, figs. 1, 2, 1917. Populus? neomexicana Knowlton, ibid., 258, pl. 53, fig. 3. Phyllites trinervis Knowlton, U. S. Geol. Surv. Prof. Paper 130, 166, pl. 24, fig. 12, 1922. There are four nearly complete specimens in the collections which resemble one or another of the somewhat variable leaves previously referred to this species. Of the specimens here shown, figure 1 resembles more closely the numerous specimens from the Laramie formation, including Lesquereux's type specimen shown in volume 7, plate 47, figure 5. The other Medicine Bow specimen, figure 2, is more of the type of Lesquereux's figure 4, and is clearly the same species as the Vermejo specimens referred to Phyllites populoides and Populusf neomexicana. Knowlton's discussions of Dombeyopsis obtusa in the reports cited above have adequately treated its taxonomic relationships and its variabihty. It may be added here that the apex of the leaf may be sHghtly emarginate, as shown in both my specimens. This species is now known from the Laramie, Trinidad, Vermejo, lower Denver, Dawson, and lower Medicine Bow formations, and is not known from beds of indisputable Paleocene or later age. I have found no modern leaves closely resembling those of this species. Analogous types are met with in the genera Ficus (Moracese) and Cocculus (Menispermacege), but these are not sufficiently similar for exact comparison. Occurrence — Corson Ranch, Wyoming, Loc. P. 372. Colledion—lJ. C. Mus. Pal., Plesiotypes Nos. 1367, 1368. Dombeyopsis trivialis Lesquereux (Plate 15, Fig. 4; Plate 16, Figs. 1, 2, 4) Dombeyopsis trivialis Lesquereux, U. S. Geol. Surv. Terr., 6th Ann. Rept. (1872), 380, 1873; Rept. U. S. Geol. Surv. Terr., vol. 7, 255, pl. 47, fig. 3, 1878; Knowlton, U. S. Geol. Surv. Prof. Paper 130, 163, pl. 13, fig. 3, pl. 14, fig. 3, 1922. From 31 specimens encountered in the field, 9 well-preserved specimens of this type of leai were collected for study. The average leaf form, shown here in figures 1 and 4, plate 16, re- sembles closely the tri-lobed figured specimen of Dombeyopsis trivialis, which Lesquereux obtained from the Laramie formation. The resemblance to the actual type specimen at the U. S. National Museum is even closer. The drawing of this specimen is not strictly correct. Actually, the central and lateral lobes are more obtuse and rounded, essentially as in tlie Medicine Bow specimens. The specimen shown here in figure 4, plate 15, was at first considered quite distinct. For- tunately, the features of the leaf shown in figure 2 seem to be intermediate in character between this smaller, more elongate leaf and the larger, more rounded leaf of figures 1 and 4. There seems also to be a gradational increase in the size of the lateral lobes, and a widening of the base from the smaller leaf form to the larger. Recognizing the types of variations to be expected in a lobed leaf form of this pattern, it seems justifiable to regard the three specimens as conspecific. In view of the variability shown it is desirable to amphfy the original description of this species, which was based on a single fragmentary specimen. FOX HILLS AND LOWER MEDICINE BOW 73 Description {Supplcjnentary) — Leavcs palmatoly trilobpd, obovate to cordate in outlinc, with 1 to 4 inconspicuous to wcll-developed subsidiary lobcs on tlie outer margins of the lateral lobes; lobation obtuse, rounded; length 6 to 10 cm., width 4 to 8.5 cm., average dimensions ap- parcntly about 8 by 7 cm.; basc broadly cuncate to cordate; secondary venation tri-pahnatc, the latcral primarics givcn off from the top of the pctiole, curving sliglitly upward and entering the latcral primary lobcs, secoiidary voins along midrib roniote, wide-angled, curving ujjward and looping near margins to join sccondarics above, 3 to 6 prominent secondaries along marginal sides of lateral primaries, the larger ones often branchcd and entering lateral lobes, the smaller ones camptodrome; tcrtiary vonation indistinct, the ncrvillos usually approaching angle of 90° to the primary aiul socondary voins; toxturo rather thin. Do7nbeyopsis trivialis has tluis far bcen reported only from the Laramie and the Medicine Bow formations. I am not aware of any other hite Cretaceous or early Tertiary occurrences of leaves which might possibly be the same. I do not Ijolieve tliat the affinitios of this fossil species aro with the existing genus Dombeya, as is inchcated in the name Dotnbeyopsis. The species of Dombeija, in my exporience, have in- variably more acutely lobed leaves with well-defined acute teeth along their margins. Although I have not seen any modern leaves which are precisely similar to the fo.ssil species, a closcr re- semblance is seen to the leaves of certain Menispermacea; : Cocculus (Cebatha) diversifolia, C. carolinum, and Menispermum canadense, and of the Vitacece, particularly Cissus rhombifolia of Mexico. Occurrence — Corson Ranch, Wyoming, Loc. P. 372; north of Walcott, Wyoming, Loc. P. 373; Craig, Colorado, Loc. P. 376. Colleclio7i—\J. C. Mus. Pal., Plesiotypes Nos. 1369, 1370, 1371, 1392. Family CAPRIFOLIACE^ Genus VIBURNUM (Tournefort) Linn6 Viburnura raarginatum Lesquereux (Plate 15, Figs. 3, 5; Plate 17, Figs. 4, 5) Viburnum marginatum Lesquereux, U. S. Geol. Surv. Terr., 6th Ann. Rept. (1872), 395, 1873; U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1873), 382, 401, 1874; idera (1874), 306, 1876; idem (1876), 510, 1878; idem, Bull., vol. 1, 380, 1875; Rept. U. S. Geol. Surv. Terr., vol. 7, 223, pl. 38, figs. 1, 4 only, 1878; Har- vard CoU. Mus. Comp. Zoology BuII., vol. 16, 51, 1888; Knowlton, Geol. Soc. Am. Bull., vol. 8, 145, 1897; Jour. Geol., vol. 19, 361, 370, 371, 1911; Berry, Canada Geol. Surv. Mem. 182, 57, 1935; Brown, U. S. Geol. Surv. Prof. Paper, raanuseript, 1937. Plalanus heerii Lesquereux. Ward, U. S. Geol. Surv., 6th Ann. Rept., 552, pl. 40, figs. 8, 9, 1886; idem, Bull. 37, 34, pl. 15, figs. 3, 4, 1887. Plalanus plalartoides (I^squereux) Knowlton (in part), U. S. Geol. Surv. Prof. Paper 101, 323, pl. 95, fig. 4, 1917; idem, Prof. Paper 130, 146, pl. 13, fig. 1, 1922. Platanus guillelma: heerii Knowlton, U. S. Geol. Surv. Prof. Paper 101, 323, pl. 96, fig. 5, pl. 97, fig. 1, pl. 98, fig. 2, 1917. Platanus marginata (Lesquereux) Heer. Knowlton, U. S. Geol. Surv. Prof. Paper 155, 81, pl. 36, figs. 2, 3, 1930. Plalanxis aceroides lalifolia Knowlton (in part), ibid., 76, pl. 33, fig. 1. Specimens referable to this species were encountered at 4 of the 6 Medicine Bow localities, although they were abundant only at LocaUty P. 372, where the field collections contained 241 specimens, of which 20 complete or nearly complete samples were collected for study. The majority of these are of precisely the same size and character as the nearly perfect original type spccimen from Black Buttes, Wyoming.i Since Lesquereux's report, this species has become known from a number of scattered localities in the late Cretaceous strata of the Rocky Mountain region. Knowlton has very ably discussed the essential distinguishing features and distribution of the species and has described fairly adequately its average leaf form.- In view of the better and more abundant material in the Medicine Bow collections it scems highly desirable to make additions and corrections to his description and to attempt to justify the inclusion of the species hsted above as conspecific with Viburnum marginatum. Description {Supplementary) — Leaves broadly ovate to obovate, narrowing abruptly above to a blunt obtusely pointed or a short acuminate tip, narrowing more gradually below to an acutely ' Lesquereux, Leo, Rept. U. S. Geol. Surv. Terr., vol. 7, pl. 3S, fig. 1, 1878. = Knowlton, F. H., L". S. Geol. Surv. Prof. Paper 155, 81, 1930. 74 CONTRIBUTIONS TO PAL^ONTOLOGY cuneate base; length 5 to 14 cm., width 5 to 11 cm., average dimensions appear to be about 9 by 8 cm.; petiole thick, long, up to 5 cm.; venation palmate, tri-nerved from the top of the petiole, strong, tlie midrib straight or slightly zigzag toward the apex, craspedodrome with 4 or 5 pairs of strong, subopposite secondaries, straight, subparallel, diverging from the midrib at an angle of 20° to 30°, more rarely up to 35° but never higher; two lower pairs of secondaries usually branched once or twice, the branches passing to the marginal teeth; lateral primaries diverging at the same angle as the secondaries on the midrib, often shghtly longer than secondaries, giving slightly trilobate character to the leaf ; each lateral primary with 4 to 6 acute secondary branches on its lower side, the lowest of which bears num.erous, usually camptodrome branches, while the upper branches are unbroken, or often once or twice branched and craspedodrome; branches occasionally also developed on upper side of lateral primaries near the margins, usually craspedo- drome; tertiary venation strong, the nervilles both broken and percurrent; margin entire below to near the middle of the leaf, becoming denticulate above, the teeth smali, usually acute, occa- sionally slightly obtuse; texture firm. Lesquereux originally included in Vihurnum inarginatum, several smaller leaves of quite different character,' which both Knowlton and Berry later agreed should be separated into a new species. The evidence from the Medicine Bow collections substantiates separation, since the smaller specimens of V . marginatum (see plate 15, figure 3) are consistently of the same character as the larger specimens. It is, of course, possible that this species may have had two entirely different leaf forms among the smaller leaves. Whether or not this is probable must remain a matter of opinion until branches bearing both kinds of smaller leaves be discovered. From the same locality from which V. marginatum was described Lesquereux obtained several larger specimens of what he called V. platanoides. These he distinguished from V. mar- ginatum by the "less numerous, more open, lateral veins, whose branches are more curved in passing up to the borders, and especially by the enlarged truncate or subtruncate base of the leaves." ^ I was at first inclined to regard the leaves of this type as large forms of V. marginatum, but failed to find a single form among the hundreds of Medicine Bow specimens which possessed the essential characters of V . platanoides. Moreover, the large specimens in the collections main- tained consi.stentIy the diagnostic characters of V. marginatum. It seems evident, therefore, that these two species are distinct, as Isjiowlton has previously concluded.' The abundant material of V. marginatum now available makes it possible to revise the status of several late Cretaceous specimens referred to various species of Platanus. The specimens which Ward obtained from the type locality of Viburnum marginatim and which he referred to the Dakota species Platanus heerii are clearly referable to the former species in every respect. The Raton and Laramie specimens referred to P. platanoides by Knowlton are easily distinguished from that species by their cuneate bases, shape, and venation. I am also convinced that the Raton specimens referred to P. guillelmw heerii and the Denver specimen referred to P. aceroides laiifolia are indistinguisliable from the larger specimens of Viburnum marginatum, which are often shghtly trilobate in shape. The previous and present record of V. marginatum indicates that it is an extremely wide- spread, abundant, and easily rccognizable species, which is limited stratigraphically to deposits of latest Cretaceous age. It is unknown from the extensive floras of indisputable Paleocene or later age of the Rocky Mountain region. It may possibly be present in the Paleocene of Texas.'' The exact formation here is in doubt, however, and the specimens (= Platanus aceroides latifolia) have not been figured, making the occurrence of doubtful value. The generic status of Viburnum marginatmn has suffered considerable vacillation between the genera Viburnum and Platanus. Lesquereux originally referred the species to Viburnum on the basis of comparison with the modern V . pubescens Pursh, V. dentatum Linn6, and V. lantanoides Michx. Heer subsequently discovered leaves which he considered conspecific with V. niarginatum in the Tertiary (Eoccne?) of Unartok, Greenland, and referred the species to Platamis. Heer's specimens are undeniably platanoid and are closely associated with typical platanoid fruits, but the leaves are clearly different in shape and basal characters from Lesquereux's specimens, and, as Holhck has pointed out, "should be regarded not only as specifically but also as generically ' Lesquereux, Leo, Rept. U. S. Geol. Surv. Terr., vol. 7, pl. 37, fig. 11, pl. 38, figs. 2, 3, 5, 1878. ^Lesquereux, Leo, U. S. Geol. and Geog. Surv. Terr., Ann. Rept. (1874), 314, 1876. ' Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 130. 146, 1922. * Berry, E. W., U. S. Geol. Surv. Prof. Paper 91, 13, 1916. FOX HILLS AND LOWER MEDICINE BOW 75 distinct." ^ Knowltoii foUowod Ilcer in refcrring the species to Platumis, but in more recent re- ports both Berry - and Brown ' have revertcd to the original reference to Viburnum, in which I concur. In comparisons with modcrn herbarium matcrial at The New York Botanical Garden, I have seen no leaves of Platanus comparable to the combination of characters observed in the leaves of Viburnum marginatum. These bear a close resemblance to several specics of Viburnum,, however, such as the species mentioned by Lesquereux, and more particularly V. opulus and V. paucijlorum of North America and V. sargentii Koehn of China. Occurrence — Corson Ranch, Wyoming, Loc. P. 372; north of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374; Craig, Colorado, Loc. P. 376. Collection—\]. C. Mus. Pal., Plesiotypes Nos. 1372, 1373, 1374, 1375. Viburnum montanum Knowlton (Plate 17, Fig. 2) Viburnum monlanum Knowlton, U. S. Geol. Surv. Bull. 163, 73, pl. 19, figs. 1, 2, 1900; idem, Prof. Paper 101, 276, pl. 52, fig. 2, 1917. Viburnumf prohlcmaticum Knowlton, U. S. Geol. Surv. BuU. 163, 71, pl. 19, fig. 4, 1900; idem, Prof. Paper 101, 276, pl. 49, fig. 9, 1917. The collections contain only 4 well-preserved specimens of this forra, of which the best is here figured. This is chflicult to distinguish from the type and figured specimens of the two species mentioned above, particularly those from the Vermejo formation. It is possible, as Knowhon has pointed out, that these specimens are variants of Viburnum marginatum Lesquereux. In the absence of gradational variants between the narrow elhptic leaves of V. montanum and the typical smaller leaves of V. marginatum shown on phite 15, figure 3, it seems necessary to maintain the former as distinct. I .suspect that a larger suite of speci- mens of this species might bring out a close rescmblance and possibly identity to the leaves referred to V. contortum,'^ which is known from the Raton and Black Buttes floras, and to V. speciosum from the Raton.^ The pre^dously recorded occurrences of V. viontanum are hmited to the Mesaverde and Vermejo formations of Colorado and to the undifTerentiated Montana group of Utah. I can suggest no better generic reference for this species than Viburnum. Occurrence — Corson Ranch, Wyoming, Loc. P. 372; Elk Mountain road, Wyoming, Loc. P. 374. Collection—V . C. Mus. Pal., Plesiotype No. 1376. DICOTYLEDON.^— POSITION UNCERTAIN Phyllites craigensis Dorf, n. sp. (Plate 18, Fig. 3) There are 4 well-preserved specimens of a very distinctive type of leaf from Locahty P. 375. I have been unable to find any fossil or living species with which a comparison could be made. It seems desirable, nevertheless, to describe and figure the species for future reference. Description — Leaves wide-elliptic, with the base broadly cuneate and the tip acutely rounded; length ranging from 8 to 13 cm., width from 4 to 8 cm., widest at the middle; petiole very stout, short, wider than the midrib; midrib very thick below, gradually thinning toward the apex; secondary venation indistinct, pinnate, 12 to 18 pairs of secondaries mainly decurrent along the midrib, diverging at an angle of about 40° near the base and a somewhat greater angle near the apex; secondaries branched irregularly and acutely, curving only shghtly upward, becoming indistinct near the margins; tertiary venation obsolete; margin entire; texture coriaceous. In both shape and venation the fossil specimens approximate the leaves of several Uving species of the genera Drimys and Illicium of the Magnohacese. The resemblance to Illicium floridanum ElUs is striking in all characters except width and expanded petiole. 1 HoUick. A., U. S. Geol. Surv. Prof. Paper 182, 166, 1936. - Berry, E. W., Canada Gcol. Surv. Mem. 182, 57, 1935. ' Brown, R. W., U. S. Geol. Surv. Prof. Paper, manuscript, 1937. < Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 101, 346, pl. 108, fig. 3, 1917. ' Knowlton, F. H., ibid.. 347, pl. 111, fig3. 1-5. 76 CONTRIBUTIONS TO PAL^ONTOLOGY Occurrence — Craig, Colorado, Loc. P. 375. Collection—U. C. Mus. Pal., Type No. 1377. Phyllites \\nlderi Dorf, n. sp. (Plate 18, Fig. 4) The single specimen figured is the only one of this kind in the coUections. I have been unable to find any figured fossil forms with which a comparison was justified. The specimen is complete enough, however, to warrant description and figure for possible future reference. Description — Leaf oval-obovate, with asymmetrical, acutely cuneate base, and short, bluntly pointed tip; length approximately 11 cm., width about 4.5 cm. at the widest portion just above the middle; midrib heavy, straight; about 8 pairs of thin secondaries, mainly opposite or subop- posite, rather eciually and distantly spaced, branching from the midrib at an angle of 50° to 60°, and converging shghtly toward the margin wliere they form inconspicuous loops with the second- aries above; margin entire; texture firm. The species is named for Mr. Beverly Wilder, who assisted in the collection of the material. Definite generic allocation of this specimen cannot be made until a larger suite of better material can be obtained. Occurrence — Elk Mountain road, Wyoming, Loc. P. 374. Colleciion—V. C. Mus. Pal., Type No. 1378. Phyllites conwajd Dorf, n. sp. (Plate 18, Fig. 2) The figured specimen is the only one of this type of leaf in the collections. It is so nearly complete, however, that it seems desirable to record its presence in the Medicine Bow flora. No comparable fossil forms have been recorded, to my knowledge. Description — Leaf obovate, narrowed rather abruptly to a slightly acuminate, rounded tip and more gradually to an apparently broadly cuneate base; length about 11 cm., width 7.5 cm. ; midrib stout and straight ; (3 to 7 pairs of prominent secondaries, occasionally decurrent, leaving the midrib at angles of 40° to 60°, arching abruptly upward well within the margin to join sec- ondaries above or their lower branches in wide loops; tertiary veins fairly prominent, widely spaced, mainly percurrent, occasionally parallel to secondaries along the midrib; margin entire; texture fairly thin. I have no suggestions regarding a generic reference for this specimen. Occurrence — Corson Ranch, Wyoming, Loc. P. 371. Collection—U. C. Mus. Pal., Type No. 1379. Phyllites colubrinoides Dorf, n. sp. (Plate 19, Figs. 4, 5) There are 8 specimens from 2 locahties which at first seem comparable to the leaves referred to Trochodendroides nebrascensis. Closer scrutiny, however, reveals a number of differences which are so constant that I am convinced that these leaves belong to a diilerent species, if not also a different genus. The main points of distinction appear to be the thick coriaceous texture, the relatively large, evenly spaced teeth, the turned-down petiole, the narrowly cuneate base and the inward-Iooping prominent secondaries and tertiaries in the leaves here referred to Phyllites colu- brinoides. Description — Leaves wide-lanceolate to elliptic, gradually narrowed to an acute apex and a narrowly cuneate base; length 3 to 6.5 cm., width 2 to 3.3 cm.; petiole turned abruptly downward, not well preserved; prominently 3 pli-nerved from the top of the petiole, the midrib and lateral primaries of almost equal strength; lateral primaries leaving the midrib at 30° angle or less, curving abruptly upward, parallehng midrib to near the apex where they fork into two veins, one of which enters a marginal tooth, the other looping inward to join the upper secondaries; one pair of prom- inent secondaries diverging acutely from above the middle of the midrib, running subparallcl to the midrib, forking into the marginal teeth, and ultimately looping inward to join the midrib near the apex; 5 or 6 prominent secondaries given off from the marginal side of the lateral primaries, looping distinctly upward to join the ones above, and giving off less prominent tertiary veins into F O X H I L L S A N D L O W E U M E D I C I N E B O W 77 the marginal tpoth; tertiary veins between midrib and lateral primaries distinrt, mainly forming upward convcx arches; margin scrrate, the teeth prominent, evenly spaced and upward-pointing; texture very coriaceous. The only describcd fossil form which I have found comparable to this species is Zizyphus meekii Lesquereux from the Ilanna formation of Carbon, Wyoming.' There arc easily recognized differences, howevcr, in shape, texture, marginal teeth, and venation, which were more apparcnt in the comparison with the actual type specimens than with Lcsqucreux's figures. The specific name given to this species impUes a close resemblance to the leaves of Colubrina (Rhamnaceae), particularly C. asiatica Brongn. and C. puhescens Kurz. The specimens are also comparable in general to leaves of Zizijphus and Paliurus of the same family. Occurrence — Corson Hanch, Wyoming, Locs. P. 371, P. 372. Collection—V. C. Mus. Pal., Type No. 1380, Paratype No. 1381. Phyllites sp. (Plate 18, Fig. 1) There are several fragments and the fairly complete specimen figured which apparently are quite distinct from any other leaf form in the coUections. I have not been able to identify this type either generically or specifically. The general shape and character of the venation suggest a reference to the genus Sti/rax. Tlie specimen is included mainly for future reference. Occurrence — Corson Ranch, Wyoming, Loc. P. 372. Collectio7i—\]. C. Mus. Pal., No. 1382. POSITION UNCERTAIN Genus PAL.EOASTER Knowlton Palseoaster inquirenda Rnowlton (Plate 19, Figs. 3, 6) Palwoaster inguirenda Knowlton, U. S. Geol. Surv. Prof. Paper 101, 278, pl. 49, figs. 5, 6, 1917. Palceoaster? similis Knowlton, U. S. Geol. Surv. Prof. Paper 130, 168, pl. 24, figs. 10, 11, 1922. The 2 specimens figured are the only representatives of this unique species in the collections. Each coasists of a whorl of segments or "leaves," apparently eight or ten in number, joined at the base and free at the apex, to form a spheroidal mass. Each segment is traversed longitudinally by a deep median furrow or rib, and perpendicularly from the midrib to the margin by closely spaced, subparallel, wavy lines. These are precisely the diagnostic features of Pala:oaster in- quirenda from the Vermejo and Raton formations. The same characters are present in the Lara- mie specimens of P. ? similis, which Knowlton maintained as distinct because of its fewer segments and its attachment to a stalk. The presence of 4 segments in the exposed half of the Laramie specimen would seem to me to imply the existence of an approximately equal number in the unex- posed half, bringing the total number of segments to 8 or more, rather than 5 or 6 as Knowlton reported. The attachment to a stalk seems equally fortuitous, and of doubtful taxonomic sig- nificance. I have therefore regarded the 2 species as synonymous. I hesitate to separate either of these species from Sterculiocarpus coloradensis Berry from the Dawson (?) formation.^ This species differs mainly in the presence of longitudinal ridges rather than furrows along the median line of the segments. I am inclined toward the view ex- pressed by Berry that remains of this type are seed capsules rather than of fohar nature. I have been unalDle to find any modern forms with which these specimens might be compared. During the past year I obtained an excellent limonitized specimen of this same unusual seed capsule (?) from the Lance formation near Glendive, Montana. The specimen was collected and sent to Professor G. L. Jepsen of Princeton University by Mr. 0. T. Lewis of Helena, Montana. Occurrence — North of Walcott, Wyoming, Loc. P. 373. Collection—lS. C. Mus. Pal., Plesiotypes Nos. 1383, 1384. ' Lesquereux, Lco. Rept. U. S. Geol. Surv. Terr., vol. 7, 275, pl. 51, figs. 10-14, 1878. ^Berry. E. W., Washington Acad. Sci. Jour., vol. 22, 119, figs. 1, 2. 1932. 78 CONTRIBUTIONS TO PAL^ONTOLOGY Carpites walcotti Dorf, n. sp. (Plate 19, Figs. 1, 2) There are 6 nearly perfect specimens from 3 localities which are obviously elongate seeds. Comparable fossil seeds are apparently rare. They bear a superficial resemblance to the seeds referred to Pisonia fructifera Berry from the Wilcox fiora ' but are clearly not the same species. Description — Seeds narrowly elliptic in fiattened view, abruptly truncate at the apex and gradually narrowed below to an acutely cuneate base; length rather consistently 1.5 cra., width 4 to 5 mm., peduncle stout, long, up to 1.8 cm., surface ribbed, the costte widely spaced and num- bering 3 or 4; apex invested with thin, short projections, the o\iter ones longer than the inner ones. These seeds are unusually complete and sliould be easily identified generically. I have as yet found no modern seeds, however, which I could consider strictly comparable. Those of Nyssa aquatica Linne, which are borne singly on long peduncles, are often of precisely the same character,^ though they may become larger and more rounded. Occurrence — North of Walcott, Wyoming, Loc. P. 373; Elk Mountain road, Wyoming, Loc. P. 374; Craig, Colorado, Loc. P. 375. Collection—H. C. Mus. Pal., Cotypes Nos. 1385, 1386. Carpites glumaeformis Lesquereux (Plate 14, Figs. 3, 4) Carpites glumceformis Lesquereux, Rept. U. S. Geol. Surv. Terr., vol. 7, 304, pl. 35, fig. 4d, pL 60, figs. 14-17, 1878. Seeds ( ?) of this type are rather abundant in the collections, being represented by niore than a dozen specimens. They are essentially lilve the type and figured specimens of this species, wliich is common in the Black Buttes flora. No other fossil occurrences are known. The exact nature of these specimens has not been determined. Occurrence — Corson Ranch, Wyoming, Loc. P. 372; north of Walcott, Wyoming, Loc. P. 373. Collection—V. C. Mus. Pal., Plesiotypes Nos. 1387, 1388. PhyUites trillioides Dorf, n. sp. (Plate 17, Fig. 3) There are but 2 incomplete specimens of this type in the collections. The general outline, venation, and heart-shaped base are typical of the leaves usually referred to Smilax. In modern leaves of Smilax, however, I have never seen forking primary veins such as are present in these fossil specimens. The leaves of Trillium, on the other hand, are almost always characterized by forking primaries (cf. T. cernuum Linne).' In the absence of more complete material, however, it is impossible to make a definite generic reference. These specimens resemble both Smilax? inquirenda Knowlton from the Laramie formation * and S. carbonensis Cockerell from the Hanna formation.^ There are enough characters preserved, however, to incUcate that they are not conspecific with either. Occurrence — Craig, Colorado, Loc. P. 375. Collection—V. C. Mus. Pal, Type No. 1389. Bracts (?) (Plate 14, Fig. 2) There are several well-defined specimens which may be seed bracts or bud scales. Although their systematic position is problematical, they are sufficiently well preserved to be recorded for possible future reference. Occurrence — Corson Ranch, Wyoming, Loc. P. 372. Collection—lJ. C. Mus. Pal., No. 1390. > Berry, E. W., U. S. Geol. Surv. Prof. Paper 156, 68, pl. 11, figs. 5, 6, 1930. ' See sheet No. 201 (,= N. uniflora Wang.) at Princeton University herbarium. ' See sheet No. 51, Princeton University herbarium. «Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 130, 118, pl. 4, fig. 5, 1922. 'Lesquereux, Leo, Rept. U. S. Geol. Surv. Terr., vol. 7, 94, pl. 9, fig. 5 ( = Smilax grandi/olia Unger), 1878. Carneoie Inst. Washington Pub. 508— Dorp Plate 1 12 Plate 2 FiG. 1—Sequoia? acuminata Lesquereux. Plesiotype. U.C. Mus. PaliEobot., No. 1295 FiGS. 2, 3 — Palmocarponf subcylindricum Lesquereux. Plesiotypes. U.C. Mus. PalEEobot., Nos. 1304, 1303 FiG. i—Zingiherites dubius Lesquereux. Plesiotype. U.C. Mus. Palseobot., No. 1306 FiG. 5—Canna cf. C.f magnifolia Knowlton. U.C. Mus. Pala^obot., No. 1305 FiG. 6—Sabalites eocenica (Lesquereux) Dorf. Plesiotype. U.C. Mus. Pala!o- bot., No. 1301 Carneoie Inst. Wasiiinqton Pub. ,')0S — Doitp Pl,ATE 2 . >5 Plate 3 FiG. 1 — Juglans leconteatm Lesquereux. Plesiotvpe. U. C. Mus. Palffiobot., No. 1310 FiG. 2—Sabalites mnntana (Knowlton) Dorf. Plcsiotype. U. C. Mus. Palseo- bot., No. 1300 FiG. Z—Sabaliles eocenica (Lesquereux) Dorf . Plesiotype. U. C. Mus. Palaeo- bot., No. 1302 FiG. 4 — Pamlanilcs corsoni Dorf . Holotype. U. C. Mus. PaliEobot., No. 1298 Car.mf.c.ik Inst. W.\shinoton' Pub. 508 -Dorp Pl.\te 3 Plate 4 FiGS. 1, 2, S—Myrica tnrreyi Lesquereux. Plesiotypes. U. C. Mus. Palaio- bot., Nos. 1309, 1307, 1308 FiG. 4 — Juglans neifberryi Knowlton. Plesiotype. U. C. Mus. Palseobot., No. 1311 FiG. 5 — Jitglans Uconteana Lesquereux. Plesiotype. U. C. Mus. PaL-Boljot., No. 1393 FiG. G—Juglans praerugnsa Knowlton. Plesiotvpe. U. C. Mus. Pateobot., No. 1312 Carnegie Inst. Wasuington 1'ub. 508 — Dorf 1'late 4 ¥-■: W'.^- \ i^^;:.i*^ % \ ■ li ^m : m ^ . • w^^ ^r ' \ ■ '~^l f 3 Plate 5 FiGS. 1, 2, 6 — DryophyUum svbfalcaturn Lesquereux. Plesiotvpes. U. C. Mus. Pateobot., Nos. 1313, 1315, 1314 FiGS. 3, 4, 5, 7 — Fict/s planicostata Lesquereux. Plesintvpes. U. C. Mus. PaLT;obot., Nos. 1320, 1319, 1391, 1321 Cahneoie Inst. Washington Pub. 508 — Dorp Pl.ATE ^'sn Plate 6 FiGS. 1, 4 — Ficus trinervis Ivnowlton. Plesiotypes. U. C. Mus. Pala;obot., Nos. 1324, 1323 FiG. 2 — Fieus tesselata Lesquereux. Plesiotype. U. C. Mus. Pateobot., No. 1326 FiGS. 3, 5, 7 — Quercus viburnifolia Lesquereux. Plesiotypes. U. C. Mus. Palffiobot., Nos. 1316, 1317 FiG. 6 — Ficus coloradensis Cockerell. Plesiotype. U. C. Mus. PaliEobot., No. 1327 Carnegie Inst. Washington Pub. 508 Dorf Tlate 6 m / m l, X^ »i3» Plate 7 FiG. 1—Ficus cowane7isis Knowlton. U. C. Mus. Palaeobot., No. 1328 . . . FiQ 2—Ficus cockcrelli Knowlton. Plesiotype. U. C. Mus. Palaeobot., No. 1322 FiG. S^Ficus crossii Ward. Plesiotype. U. C. Mus. Palajobot., No. 1325 . . FiG. i—Laurus socialis Lesquereux? Plesiotype? U. C. Mus. Palsobot., No. 1335 FiG 5—Laurophyllum meeki Dorf. Plesiotype. U. C. Mus. Palaobot., No. 1334 Caknegie Inst. Wasiiington Pun. 508 -Doiif Plate 7 k< X \ ';v-^< V/\^^"AiP^."/5T Plate 8 FiG. l—Crcdneria protophylloides Kjiowlton. X %. Plesiotype. U. C. Mus. Paloeobot., No. 1329 Carnegie Inst. Washington Pub. 508— Dorf Plate 8 Plate 9 FiG. l^Citinamomum linifolium Knowlton. Plesiotype. U. S. Nat. Mus. Coll., No. 1331 ." FiG. 2—Pisonia? racemasa Lesquereux. Plesiotype. U. C. Mus. Palfeobot., No. 1330 FiGS. 3, i—Cinnamomum affine Lesquereux. Plesiotypes. U. C. Mus. Palffiobot., Nos. 1332, 1333 FiG. b—MagnoKophtjUum cordatum Dorf. Cotype. U. C. Mus. Pateobot., No. 1345 Carnegie Inst. Washington Pub. 508 — Dorf Plate 9 Plate 10 FiG. 1 — Magnoliophyllum cordatum Dorf. Cotype. U. C. Mus. Pala>oliot., No. 1344 FiG. 2 — Magnnlia lakesii Knowlton. Plesiotype. U. C. Mus. Palaeobot., No. 1340 FiG. 3 — Magnolia dakotana Berry. Plesiotype. U. C. Mus. Palseobot., No. 1341 FiG. 4 — Magnolia nervosa (Knowlton) Brown. Plesiotype. U. C. Mus. Palajobot., No. 1342 CaRNEQIE InST. WASHmOTON PuB. 508 — DORF Plate 10 ^^^k '..v^ --> /• /^: Plate 11 FiGS. 1, 4, 6, 7 — Trochodendroides nebrascensis (Newberrv) Dorf. Plcsio- types. U. C. Mus. PaliBobot., Nos. 1337, 1336, 1339, 1338 .... FiGS. 2, 3, 5 — Leguminosites arachioides minor Berry. Plesiotypes. U. C. Mus. Pateobot., Nos. 1348, 1346, 1347 FiG. S—Magnolia pulchra Ward. Plesiotype. U. C. Mus. Palaobot., No. 1343 Carnegie Inst. Washinuton 1'ub. 508 Dokf Plate 11 t\ f *■ -j» X- ■ - --^■r •>".■ -■«rr""t ,%^.-^ ■ J ..j^t Jij — Plate 12 FiGS. 1, 2, 3 — Celaslrus taurinensis Ward. Plesintypes. U. C. Mns. Palaeo- bot., Nos. 1350, 1352, 1351 FiGS. 4, 5 — Rhamnus salicifoluts Lesquereux. Plesiotypes. U. C. Mus. Palaiobot., Nos. 1353, 1354 FiG. 6 — Zizyphiis hendersoni Kjiowlton. Plesiotvpe. U. C. Mus. Palaiobot., No. 135S ' Carnegie Inst. Washinqton Pub. 508 — Dorf Plate 12 ►■■ / Plate 13 FiGS. 1, 5, 6 — Grewiopsis saporian-a Lesquereux. Plesiotypes. U. C. Mus. Palaeobot., Nos. 1364, 1363, 1365 FiGS. 2, 3 — Cissites lobahis Dorf. U. C. Mus. Palaeobot., Type No. 1359; Paratype No. 1360 FiG. 4 — Rhamnus salicifoUus Lesquereux. Plesiotype. U. C. Mus. Pakeobot., No. 1355 Carnegib Inst. Washington Pub. 508 — Douf Plate 13 ' / ,'. Plate 14 FiG. 1 — Apeibopsis ? laramiensis Ivnowlton. Plesiotype. U. C. Mus. Palaeo- bot., No. 1362 FiG. 2— Bracts (?) U. C. Mus. Palajobot., No. 1390 FiGS. 3, 4 — Carpites glumaeformis Lesquereux. Plesiotypes. U. C. Mus. Palaeobot., Nos. 1387, 1388 FiGS. 5, 6 — Rhammis deburni Lesquereux. Plesiotypes. U. C. Mus. Palaeo- bot., No. 1357 CaRNEGIE InST. WASIlINliTON VvB. 508 — DollF / .V Plate 14 ■ !*i- I 9 3A Plate 15 FiQS. 1, 2 — Domheyopsis obtusa Lesquereux. Plesiotypes. U. C. Mus. Palseobot., Nos. 1367, 1368 FiGS. 3, 5 — Vihumum marginatum Lesquereux. Plesiotypes. U. C. Mus. PalEeobot., Nos. 1373, 1372 FiQ. 4 — Domheyopsis trivialis Lesquereux. Plesiotype. U. C. Mus. Palieo- bot., No. 1369 Carneqie Inst. Washington PnB. 508— Dohf Plate 15 ■s -*«(»., ■..2] Plate 16 FiGS. 1, 2, 4 — Dombeynpsis trivialis Lesquereux. Plesiotypes. U. C. Mus. Palaobot., Nos. 1370, 1371, 1392 FiG. 3—Cornophyllum wardii Dorf. Type. U. C. Mus. Palaobot., No. 1366 Carneqie Inst. Washington Pub. 508— Dorf Tlate 16 ■-;^ i y Plate 17 FiG. 1 — Viiis slanloni (Knowlton) Brown. Plesiotype. U. C. Mus. Paliro- bot., No. 1361 FiG. 2 — Vihurnxim montanum Knowlton. Plesiotype. U. C. Mus. Palceo- bot., No. 1376 FiQ. 3—Phyllites trillioides Dorf. Type. U. C. Mus. Palceobot., No. 1389 . . FlGS. 4, 5 — Vibumum marqinalum Lesquereux. Plesiotypes. U. C. Mus. Palaeobot., Nos. 1374, 1375 (^AnNEGiE Inst. Washinoton Pub. 508 DORP Platb 17 Plate 18 FiG. 1—PhylHtes sp. U. C. Mus. Palxobot., No. 1382 FiG. 2—Phyllites conwayi Dorf . T^-pe. U. C. Mus. Palseobot., No. 1379 . Fia. Z—Phyllites craigensis Dorf . T^-pe. U. C. Mus. Palaeobot., No. 1377 FiG. ^—Phyllites wilderi Dorf . Type. U. C. Mus. Pateobot., No. 1378 . Carnegie Inst. Wasiiinoton 1'ub. 508 — Dorp Plate 18 »IP.1 Plate 19 FiGS. 1, 2—CarpUes umlcotti Dorf. Cotvpes. U. C. Mus. Palseobot., Nos. 1386, 1385 FiGS. 3, 6 — Palaeoaster inquirenda Knowlton. Plesiotypes. U. C. Mus. PalEeobot., Nos. 1384, 1383 FiGS. 4, 5—PhylKtes coluhrinnides Dorf. U. C. Mus. Palaeobot., Type No. 1380; Paratype No. 1381 FiG. 7—Pistacia eriensis Knowlton. Plesiotvpe. U. C. Mus. Palaeobot., No. 1349 Carneoie Inst. Washixgton Pub. 508— Dorf Plate 19 OE924 D6b 1938 V.1 Dorl, Erlmq/UpperCrei ceousfloras i i: I rihiliri fli «iiiiiii.u iiiriiiir" i 3 'jI 1 8768