DEPARTMENT OF THE INTERIOR HUBERT WORK~ Secretary.
UNITED STATES GEOLOGICAL SURVEY GEORGE OTIS SMITH, Director
Professional Paper 136
THE FLO~A OF THE RIPLEY FO.RMATION
BY
EDWARD WILVER BERRY
W A.SHINGTON GOVERNMENT PRINTING OFFICE 1925 ADDITIONAL COPIES OF THIS PUBLICATION MAY BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT 50 CENTS PER COPY CONTENTS
Page Introduction ______1 General relations of the Cretaceous deposits ______1 The Ripley formation ______--·-- ______·______3 Historical sketch ______.. ______3 Areal distribution ______~ ______3 Lithologic character _____ .______3 Stratigraphic relations ______~ ______3 Subdivi~ons------4 The Ripley flora------~------~------ 5 Occurrence------5 Composition, origin, and evolution ______5 D~tribution ______7 Analysis of other Upper Cretaceous floras ______..: ______15 General considerations ______15 Eutawflora------~------~------~------18 Black Creek flora------~------19 Magothy flora ______.______19 Patoot flora ______20 Verinejoflora ______~------~------21 Vancouver Island ______.:. ______21 Colorado grouP------21 Montana group ______- __ ·_- ______·__ ---______21 Japan------~------22 Portugal ______· 22 GermanY------ 22 Gosau flora------23 France ______~------~------23 23 Syst!l~:~~~gd~:cr~~~~d~:t~~-~~-~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~-~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 25 IndeX------93
ILLUSTRATIONS
Page PLATES I-XXIII. Upper Cretaceous plants from the Ripley formation ______92 FIGun.E 1. Diagrammatic section of the Cretaceous from Tennessee to Georgia, showing the time rela tions of the different stratigraphic units and their relation to underlying and overlying forinations ______4 2. Map of North America showing the area of outcrop of the Ripley formation and the generalized relations of land and sea during Ripley time------·----~------14 3. Sketch map of the world showing the location of fossil floras that have been compared with the Ripley flora ______18 4. Restoration of Geonomites visianii Berry ______----~--- ____ ·______38 5. Sketch map of the world summarizing the geologic history of Liriodendron ______55 6. Sketch map of the world. summarizing the geologic history of Zizyphus ______69 III 1.,HE FLORA OF THE RIPLEY FORMATION
By EnivARD W. BERRY
INTRODUCTION of the field work and ventured to predict that more detailed work would res'!Jlt in the In 1919 the Geological Survey published an discovery of additional fossiliferous outcrops. extensive paper by me entitled "Upper Cre The present contribution serves to fulfil this taceous floras of the eastern Gulf region prediction in part and to supply the first de in Tennessee, Mississippi, . Alabama, and tailed information regarding the flora of that .Georgia." 1 Despite this .comprehensive title part of the Upper Cretaceous section of which the work was. primarily an account of the our ignorance was most ·profound. . Both l;>io large flora found in the Tuscaloosa formation, logically and geologically this information is the earliest Upper Cretaceous formation in of the greatest importance, because it fur that region. Between the date when the nishes a measure of the rate of transition manuscript had been transmitted, in 1912, and between the Cretaceoi1s and Eocene floras, as the time of publication scattered collections well as data for. correlation and distribution. from the Eutaw and Ripley formations had Spec~al acknowledgment is due to Bruce been made, and an account of these was in vVade for the industry and skill with which cluded in the report. Although of consider,.. he has studied the Cretaceous of Tennessee able interest as the only known land plants of ·and collected the material upon. which the the Eutaw and Ripley, these plants were few: present paper is so largely based. The types and not especially well preserved, as I had are the property of the Johns Hopkins Uni not had the good fortune to encounter in these versity, but they will be depos:ted in the formations other than very carbonaceous leaf United States National Museum as soon as bearing clays, with drifted and consequently the studies in hand are completed. more or Jess broken materials. It was realized that better-preserved and more representative GENERAL RELATIONS OF THE CRETACEOUS DEPOSITS material might be found in clay lenses in the upper embayment, but neither Stephenson nor The region discussed in this paper includes I had been successful in locating .such out that part of the Atlantic Coastal Plain lying crops, and it was not until Wade's very de south and west of the southern Appalachian tailed county studies for the Tennessee Geo province and east of Mississippi River. logical Survey that this gap in the paleobo The area of outcrop of the Upper Cretaceous tanic history of the Mississippi embayment deposits borders the inland margin of the was satisfactorily filled. All the forms known Coastal Plain, and the belt from which deter from the Ripley are included in the prese.nt minable plant fossils have been collected ex contribution. The bulk of the materials have tends from the vicinity of the Kentucky come from Henry and Carroll counties, in boundary, in west-central Tennessee, .south western Tennessee-a region that has also ward and eastward .to west-central Georgia. furnished the largest single florule of lower · Both Lower ( ~) and Upper Cretaceous de Eocene CVVilcox) age. . posits are represented in this area, but· the In the introduction to my account of the former are much less extensive and varied tJ pper Cretaceous floras of the eastern Gulf than in tlw· middle Atlantic slope, to the region I emphasized the reconnaissance nature northeast, or in the western Gulf area, to the west, in Texas and Mexico. Supposed Lower 1 U. S. Geol.. Survey P1·of. Paper 112, 1919. Cretaceous sediments are present only in the 1 2 THE FLORA OF T'HE RIPLEY FORMATION
extreme eastern part of the area, where the of Alabama it overlies the Tuscaloosa without southwesterly extension of the Potomac group apparent unconformity. Beyond the Tusca of the ~fiddb Atlantic States has been consid loosa region in Tennessee the Eutaw consti ered to cross Georgia along the fall line as an tutes the basal formation of the Coast•a tn Berry, ill. W., The age of the supposed Lower Creta<'eous down in moderately shallow wat~r, as is of Alabama: Washington Acad. Sci. Jour., vol.. 13, pp. 433-435, figs. 1-6, Dec. 4, 1923. 2 U. S. Geol. Survey Prof. Paper 112, p .. 26, 19·19. THE RIPLEY FORMATION 3 shown by the abundance of fossil Ostreidae, erature containing absolutely no re:£erence to and seems to owe its peculiar character to the fossil plants except brief preliminary papers 4 absence of terrigenous . materials. It is a and the description of Ripley plants in Pro lithologic and not a chronologie unit, the care fessional Paper 112. ful faunal studies of. Stephml.son having shown AREAL DISTRIBUTION · that it is partly the equivalent of the Eutaw and largely the equivalent of the Ripley. The The beds to which the name Ripley foni1ation various faunal lines have been traced success has been applied crop out from Twiggs County, fully frmn the ·sand and clay of the Chattahoo in central Georgia, to central Alabama, where chee basin through the chalk to the sand and they are interrupted by the synchronous de clay of northeastern Mississippi. The bear posits of the upper part of the Selma chalk. ing of the Selm.a chalk on the Upper 'I'he Selma chalk forms the top of the Upper Cretaceous physiography is important, be Cretaceous for most of the distance between cause the absence of run-off carrying terrig central Alabama and Houston County, Miss. enous materials that made possible the for Fron1 that point to the head of the Missis mation of the chalk denotes a downwarping sippi embayment the Ripley .deposits are con of the upper embayment and a shifting north tinuous. In Georgia the outcrop is from 10 ward of the stream deltas that existed during to 15 miles wide; in Alabama to the line where Tusca.loosa time. it passes into the chalk phase it is frmn 25 The deposits of the Ripley sea extend fronl to 30 miles wide; and beyond the chalk in Twiggs County, in central Georgia, to the northeastern ~1ississippi it narrows to about head of the Mississippi embayment, in south., 12 miles at the 1\'Iississippi-Tennessee bound ern Illinois, and probably westward through ary. It is about 6 miles wide at the Ten:nes the western Gulf region, although there is a see-ICentucky boundary and maintains that wide interval in Missouri and Arkansas that width across ICentucky and southern Illinois. was transgressed by Eocene deposits,. large iueas of which were subsequently removed by LITHOLOGIC CHA:i.ACTER the action of ~1ississippi River during l~te · The Ripley materials are predominantly Tertiary and Quaternary time. massively bedded dark-gray to greenish-black These deposits of the Ripley sea do not con marine sands, more or less calcareous, mica stitute a single lithologic or forniational unit, ceous, pyritiferous, glauconitic, and argilla because in the eastern Gulf region those in the ceous, according to the conditions of sedimen Chattahoochee basin are separated from those tation in different parts of the Ripley sea. in northeastern Mississippi .and the States They are usually fossiliferous, in some places farther north by the western Alabama area, enough so to constitute shell marls. Ledges in which the Selma chalk continues to the top of sandy limestone are developed locally. of the Upper Cretaceous. Some distance east of the Selma chalk, as in the Chattahoochee basin, the Ripley has been THE RIPLEY FORMATION estimated (probably overestimated) to be HISTORICAL SKETCH about 1,000 feet thick, but it thins eastward in Georgia, where it becomes ·more definitely The Ripley formation was named in 1860 a deposit of littoral sands. .It is also thinner by IIilgard 3 after the town of Ripley, in in the upper embayment, having a reported Tippah County, .~1iss., which is 3 miles south thickness of 400 feet at 'Vickliffe, ICy. ; 54 west of the classic fossiliferous Ripley deposits feet at Ca.iro, Ill.; and 204 feet at. Paducah, exposed along Owl Creek. Until the studies. Ky., as estimated from well records. by Stephenson, previously mentioned, the lim its of the Ripley had been indefinite for more STRATIGRAPHIC RELATIONS than half a century, any fossiliferous Upper·. Cretaceous sand having usually been consid The Ripley forni.ation rests without appar ent unconformity on the . somewhat similar ered Ripley. The history of the study of th~se deposits need not concern us further; the lit- 'Berry, E. W. Contributions to tbe· Mesozoic flora of the Atlantic Coastal Plain-VI. Georgia: Torrey Bot. Club Bull., a Hilgard, El. W., Report on the geology and agl'iculture of vol. 37, pp·. 503-511, 1910; XI, Tennessee: Idem, vol. 43, pp. Mississippi. pp. 62,. 83-95, 1860. 283-304, pl. 16, 1916. 4 THE FLORA OF THE RIPLEY FORMATION $ands of the upper part of the Eutaw in the dence sand member of the Ripley comprises Chattahoochee basin. Eastward in Georgia it irregularly bedded sand and clay of shallow lies unconformably on the supposed Lower marine and littoral origin which can be traced Cretaceous. Westward across parts of Ala from Crenshaw . County, Ala., to Twiggs bama and Mississippi it is developed above the County, Ga. Like the Cusseta sand it thick chalk phase (Selma chalk), which, changing ens toward the northeast, and it is indis gradually northward into a clay~· underlies Lnguishable from the Cusseta in the region it as far as the southern boundary of l(entucky. where the tniddle. member of ~he Ripley has. Fr01n this point to the head of the embayment disappeared. It continues to the top of the it rests on the ~mbayment floor of Paleozoic Cretaceous in that region. rocks. It is overlain throughout its extent The Ripley deposits of western Tennessee by the unconformable deposits of the Eocene, comprise irregularly bedded sand, clay, and although a limestone at the base of the Mid marl which sinqe the days of Safford (1864) way group (basal Eocene) at Owl Creek, Miss., have been referred to the. Ripley formation. which Lowe correlates with the Clayton lime Stephenson, in 1914, proposed the name ~1c. stone of Alabama, was included in the Ripley N airy sand member for these beds. The &e. by Hilgard. tailed work of Wade has shown that the H.i p- TENNESSEE MISSISSIPPI ALABAMA GEORGIA Eocen c -id~ .e-E --_S_e-lm_· O::L ~:;:~btf-~>- • ~-~alk- cE FIGURE 1.-Diagrammatic section of the Cretaceous from Tennessee to Georgia, showing the time relations of the different stratigraphic units and their relation to underlying and ove~lying formations \il Ill:;.~ m~ ...~l SUBDIVISIONS ley of the upper embayment was predomi nantly of . marine origin and that several Disregarding earlier interpretations we horizons can be discriminated. These from may note that in the Chattahoochee area and eastward in Georgia, Stephenson segregated the oldest to the youngest are the Coon Creek, the Ripley into three members-a lower Mr;Nairy. sand, and Owl Creek. The 1nost 'named the Cusseta sand member, a middle extensive marine fauna yet found in the Rip unnamed member, and an upper or Provi ley comes from Coon Creek, Tenn., and ·the dence sand member. The Cusseta sand mem second fauna in size and variety is that of ber comprises 200 ·to 300 feet of irregularly Owl Creek, Miss., at the top of the formation. bedded, mostly nonglauconitic sand with sub The intervening beds are less· fossiliferous, ordinate clay lenses of shallow marine origin, but certain clay lenses that mark a horizon in becoming thicker and more littoral in cha.r- the MeN airy sand as limited by 'Vade con . acter toward the northeast as the more typi tain an abundance of fossil plants and. are cally marine beds of the middle member pinch interpreted as representing deposition 111 out. The · Cusseta sand member contains a .·coastal lagoons. . few ·fossil plants which indicate that it repre The chronologie relations of the Upper Cre sents the whole or a part o:f the time interval taceous formations of the eastern Gulf area of the Coffee. sand member of the Eutaw as interpreted by me are shown diagrammati formation in wester·n Tennessee. The Pro vi- cally in Figure 1. OCCURRENCE, COMPOSITION, ORIGIN, AND EVOLUTION 5 THE RIPLEY FLORA tonites, Geinitzia, Potamogeton, Alisrnaphyl lum, Dr7jophyllu-m, F'ag'lts, Oeltis, Oedrela, OCCURRENCE Acer, Dillenites, Oinnamomwm, N ectandra, Determinable fossil plants have been found Myrcia, Eugenia, Oo'rnophyllum, Ohrysophyl at 13 different .localities that .are referred to lum, and Acer(l)tes, which have not been found the Ripley formation. Those in Alabama and at the Perry place. Georgia were described in Professional Paper ~1oreover, in the larger genera, such as 112 (1919), already cited. ·Those discovered llfyrica, with 11 species, only one of these is in western Tennessee are due almost entirely common to the two localities; only two of the to the field work of Bruce ·vv ade, and the Leguminosae are common to both localities; localities, with the exception of the Big Cut 1n of the seven species of Oelastrophyllu1n b~t ~1cNairy County,have not been visited by me; one is found at both outcrops; of the nine figs the stratigraphic description of these localities none are common to both; of the five species .will be omitted from this report, as the Ten- of Apocynophyllum none are common; of the nessee Geological Survey is publishing a series three species of Zizyphus only one is com of county repotts that will contain detailed mon, and that one is r'are at one locality and discussions of these plant-bearing outcrops. abundant at the other; and of the three species The localities in Alabama and Georgia fur of Te1·nstroemites only one is found at . both nish but few and on the whole not well-pre localities. served fossil plants. They are separated by As these two localities have been shown by some 350 miles of coast line of the Ripley sea the mapping to represent practically the same from the southernmost locality for Ripley horizon these marked differences in apparent plants in Tennessee, and this in turn is about facies can be due to but two causes-th'at is, 100 miles south oi the highly fossiliferous lo they are to be accounted for solely as the re calities in Henry and Carroll counties. At the sult of accidents of preservation or discovery ·other Tennessee localities· the remains in char or as being inherent differences due to local acter and condition of preservation are 1-r{~lcl; differences in environmental conditions, at the like those of Alabama and Georgia. It is at time of growth. Possibly a combination of the Perry place, in Henry County, and the these two explanations is the correct solution. Cooper pit, in Carroll County, that most of I am indined, however, to think that in this the information regarding the Ripley flora as in many other fossil fl Those that were less sheltered were filled with · caloosa time, as I have indicated in the discus argillaceous sand such as occurs' west of Camden sjon of Tuscaloosa history. 5 The Ripley main and farther south, where most of the plant land appears to have been well wooded. fragments were broken up and the more or less There is no evidence of distinctively palustrine carbonaceous clays contain identifiable frag conditions at the two principal plant localities, ments of only the more resistant types: The although elsewhere in deposits of this age lagoons represented by the Perry and Cooper such evidence is present. Se·veral supposed localities were more sheltered and protected· aquatic plants are included in the collection, from wave action, at least for long intervals, however, and the number of coastal types is during which the 'fine muds were deposited. considerable, as I have shown in the analysis That there were no considerable streams of the flora. emptying into the Ripley sea in the Tennessee DISTRIBUTION region is rendered very probable by the history of Tennessee River, to the east. The Tennes The following systematically arranged table see appears to have been the master stream at shows the distribution of the Ripley species at that time, following much the same course that the several localities of this age from which it does at present but farther east, its distribu identifiable plants have been collected: taries having migrated northward during Tus- 5 U. S. Geol. Survey Prof. Paper 112, 1919. Distribution of the Rip~ey flora Tennessee Alabama Georgia :>. l=l :>. :>. l=l l=l 0 t' .!:I s s ~ 8 ~ ~ ·a 0 0 ·~ .s ::l ;; ~ ~ § .0 :g 0 ~ ~ "' 0 -e 0 .c: z z 0 0 0 0 0 ::~ P=l <:.> c:J 0 a ~"' o::l ~ P=l P=l"' P:l ~ ~0 ~ :>. "'~ t' 0 1E» :>. gJo ~» .!:I :>. ~» ~» :>. l=l l=l~ t' t' .!.:~ ~ ~:>. ~=:~§ P:l ~ >§ .:t§ ,c: .... ~§ ~§ z"' ~§ ~§ 0 ig <:.> Phylum PTERIDOPHYTA Order FILICALES Aspleniltm calopteris (Debey and Ettings- hausen) Heer ______X ______---- __ - _____ - ___ ---- ____ ------Raphaelia neuropteroides Debey and Ettings- hausen ______X X ______- _------__ ------Raphaelia minuta Berry______X ______------____ ------Monheimia aquisgranensis Debey and Ettings- hausen ______._ X ______------Taeniopteris sp ______X X. ______---- - _-- ____ - ___ ---- ____ ------Dryopterites stephensoni Berry ______---- ___ ~ ____ - ___ ---- ____ ---- X Phylum CoNIFEROPHYTA Protophyllocladus lobatus Berry______X ______---- ______---- ____ ------Widdringtonites reichii (Ettingshausen) Heer_ ____ X X ______---- ____ ---"' ____ ---- ____ ------Moriconia cyclotoxon Debey and Ettings- GeinitziaM~:i~~~~~ formosa ~~e-;i~;~~ Heer --se;;:Y ______~======~ _-~ X _ =______======----= = = =____= = = ------======Cunninghamites elegans (Corda) Endlicher ___ ~---- ____ ------.------X Araucaria bladenensis Berry ______X ---- Araucaria jeffreyi Berry ______X Dammara acicularis Knowlton______X ______~ 8 THE FLORA OF THE RIPLEY FORMATION Distribution of the Ripley jlo1·a-Continued Tennessee Alabama Georgia Phylum ANGIOSPERMOPHYTA Class MoNOCOTYLEDONAE Order N AIADALES Family N:AIADACEAE Potamogeton ripleyensis Berry ______X ------Alismaphyllum cretaceum Berry ______·______X Order LILIALES Family DIOSCOREACEAE Dioscorites cretace us Berry ______~ ______X ______- _------_------Doryanthites cretacea Berry ______-.--- ______---- ____ X ___ _ Order ARECALES Family ARECACEAE Geonomites schimperi Lesquereux ______X ____ ---- X ______X ------Sabalites SP------______X ____ X ____ ---- ______ Class DICoTYLEDONAE Series CHORIPETALAE Order JuGLANDALES Family JuoLANDACEAE Juglans similis Knowlton ______X J uglans tennesseensis Berry ______J uglans wadii Berry ______..; ______X ======~======Order M YRICALES Family M YRICACEAE . 1 ~~~i~: ;:~ii !~~t,.-B~~;y~::~:::::::::::~:: ~ :::: :::: :::: ::::!::~~ :::: ::~~ ~::: :::: ~::: :::~ :::: Myrica ripleyensis Berry ______!.· ______.:. X ______-: _ X ____ ---- ______Myrica cooperensis Berry ______---- X ·- _____ -- ___ ------·------.------Myrica johnstrupi (Heer) Berry ______X ______------Myrica minor Berry ______.___ X ______.______------Myrica ornata Berry______X ______------Myrica torreyi Lesquereux ______;______X ______------:---- ___ .:. ·------Myrica torreyi obtusata Berry______X ______------Myrica brittoniana Berry ______X ______Myrica brittoniana obtusata Berry ______X ""'--- ______------Order SALICALES Family SALICACEAE Salix gardneri Knowlton? ______I X DISTRIBUTION 9 Distributrion of the Ripley jl01·a--Continued 1'ennessee Alabama Georgia » 5:l » 5:l 5:l t' 0 t' :::· 13 a 5:l § ·o; 1:l 0 0 .s 1:l ~<1) ~ ·a .0 ·~ ::i ::i 0::1 1:l<1) ::i 0::1 :g 0 0 .0 Ot- z z 0 -e 0 0 0 0 ~ <.> a o:l ;:g 0 o§ ;:g 0 p::: ~ ~ d ..... 0 ~ :::- :a » gjO » :::- ~ -» » it>. » <1) ~a !:l ..... t' ·a ~-t- ~ ~» -!:l .$1:l <1)~ -~ >§ .8§ ~§ ~::i ~§ z ::i::i ~§ a§ 0 ceo !Eo :S~ <1lo <.> ceo ~s aS .... Ql so so ceo ~0 <.> .... g)O ~0 g~ o::~O ~0 Q)o ~ 'aO oo ~0 »O 0::1 0 Ql ~ ·a ::i 0 UJ ~ ::i 'a ~ gj ~- ~ 8 ~ 1il -Ql ~ p. .... 5:l .... 0 .... c 0 0::1 '§'0 0::1 0::1 -~ a te 0 bll te a 1il 0 Ql Ql .., Q 0 Ql 0 Ql <1) p., 0 z z ..... z 0 ~ z 0 z z ------1------z Order FA GALES Fa~ily FAGACEAE I Dryophyllum graCile Debey __ ------____ X X X ------.. ------Dryophyllum protofagus Berry______X Fagus ripleyensis Berry .. ______X Order UnTICALES Family ULMACEAE Celtis cretacea Berry ______.:. _____ X ---- ______---~ ______ Family MonACEAI l!mFicus ~f~i~!~~~fi.~~r=~~~~~~~~~~~~=~~~~~~~cooperensis Berry______}= _ _ _ =~r~~ X ______~~~~ ~~~= =~~~ =~~~ ~~~~ ~~~= ~~~~ ~~~~ ~~~~ ~~~~ _ ~~~~:~ ~~~:-~A~~~~i~B~~-"Y~~======--~- --x- ======-======:::: -~- =~,:-:: Order PLATANALES Family PLATANACEAE Platanus ripleyensis'Berry ______.______---- ______-----.--- ___ .: ____ X -~-- ______.Platanus SP------'------.---- ______X _____ :.. __ Order RAN ALI Family MAGNOLIACI Liriodet~dron lar~~iense Ward ______X ____ !____ ------1------Magnolia capellnm Heer ______! ______X ______. I . . Family MI Family CAPPAUIDACI Distribution of the Rip~ey flora-Continued Tennessee Alabama I Georgia p.. ... s:l s:l s:l 0 C.' C.' ::; ::;· 0 a a 0 l:l 8 ~ ste ·a ·a s:l 0 ::l § ~ l:l ::l .c .c -~ tl ~ ::l 0 0 a .c o_e. 0 0 jl:l z<:.> z<:.> til a 0 0 0 0 0 jl:l jl:l ~ p.. '0§ ~ d' ""'0 ~ ~ d' p.. gso ·» p.. p.. ·» ... p.. p.. t:l 8 ~~ t· -~ ,!C ... ~ a >§ .a§ fl: p.. ti31:l z .:i§ ~s:l ~§ a§ ~ 0 ..0 t:l '0 ::l . i§ i§ <:.> teO !So ~0 so go ~5 teO 8o s:lO .... o oo s:lO »0 ~ ....- ~ ~~I §8 030 ~ ::;o ~0 ~ jl:l d ::l 0 UJ r"'l ·a s:l 0 ~ ::l 'a ... ·jl:l gs s:l- I ~ ~ jl:l ~ ,.!C 0. s:l 0 ... t' 0 a a s~ a 0 a a -~ d a ai 0 ~ ~ ~ 0 ~ .!:!! ~ 0 ~ ~ P; 0 z z .... z 0 z jl:l z 0 z z ------Family CA;ESALPINIACEAE -,- Caesalpinites perryensis Berry ______X ____ ---- ____ ------Caesalpinites ripleyensis Berry______X ____ ---- ______--- _____ ------Bauhinia rip;~~:~·p::::~~~~~~: ______-~- -~ ______X ______X _---1--- _ Dalbergia cretacea Berry ______~_ X ------Dalbergia perry ana Berry ______X Dalbergia prewilcoxiana Berry ______X LEGUMINOSAE (position uncertain) Leguminosites cariavalioides Berry ______X X ______---- ____ ------Leguminosites perryensis Berry ______X ______---- ______---- ____ ---- Leguminosites carrollensis Berry ______---- X ------1------Gleditsiophyllum preovatum Berry______X ______------______-- _------Gleditsiophyllum proeocenicum Berry______X ______.. _ ---- ______-- ___ ------Gleditsiophyllum aristatum Berry______X _--- _____ - _- ----' _------Order GERANIALES Family MELIACEAE Cedrela prewilcoxiana Berry______X _ -- - ______- ---- ____ ------ ·Family EuPHORBIACEAE Euphorbiophyllum petiolatum Berry______X X ---- X X ---- Euphorbiophyllum tennesseensis Berry______X _____ - ______--- _____ ------Euphorbiopl!yllum cretaceum Berry ______X ______------____ ------Eupho-:biophyllum antiquum Saporta and . Marion______X _--- _____ - __ --- _____ ------Manihotites georgiana Berry ______-:______X -- ______---- X Order SAPINDALES Family CELASTRACEAE Celastrophy llum carolinensis Berry ______X ------ Celastrophyllum ripleyanum Berry______X X ------7 ------ Celastrophyllum minimum Berry______X ------I ------.--- Celastrophyllum cassinoides Berry______X Celastrophyllum perryi Berry______X Celastrophyllum variabilis Berry______X Pachystima? cretacea Berry ______.:.______X Family AcERACEAE Acer cretaceum Be:rry ______- ______X _ -:.. - - r------Order RHAMN ALES \ I Family RiiAMN ACEAE . . I ~~~fff€E!~~[fifr-~=~~~~~~~~~=~~~~ -~~-~: ~~~~ ~~~t~~ ~j~~~~ ~~~~· ==~= ~~== ~~~~ ~~~~ ~~~~ . DISTRIBUTION 11 Distrib1ttioo of the Rip~ey fior~Continued. Tennessee Alabam.a Georgia ...... ::l l:l >. >. ~ ...>. t' t> ~ :::3 8 ·c; ::l 0 c s 0 0 c: Cissites crispus Velenovsky ______. __ 1 ______X ______Cissites panduratus Knowlton _____ -----_____ X ------Order MALVALES Family TILIACEAE Grewiopsis ripleyensis Berry""' ______·_--- ______-- _ X ------Grewiopsis inequilateralis Berry _____ ---_____ X X Family STEUCULIACEAE Sterculia suowii tennesseeusis Berry----- ______------X Order P ARIJDT ALES Family DrLLENIACEAE Dillenites cretaceus Berry ______X ------ Family TERNSTROEMIACEAE Ternstroe1'nites ripleyensis Berry ______X. ____ ------Ternstroemites ten nesseensis Berry __ ---- ______X ------ Ternstroemites cretaceus Berry ______---_____ X X ----1------. Order THYMELEALES Family LAURACEAE Cinnamomum newberryi Berry ______X. X ------X ----~------· Cinuamomum newberryi lanceolatum Berry ______X ____ ------____ ------Cinnamomumnewberryiminimum Berrv ______X ____ ---- ____ ---- ·------'------b . 11· t" B ~ · X I c~nnamomum. n~w erry1 e_1p 1cum ~rry______------____ ------~------Cnu.lamomum J?Iaespectabile Berry ______X ------~------Nectandra prohfica Berry ______·---- X ------..,------Malapoenna horrellensis Berry ______..:_ -· ______-- - ______---- X X ------~~~~: ~t~!~~~r.:~f;~~:::::::::::~::::::~ s:~: :::::::::::::::: ~::: :::= :::r:= :::: :::: :::: Laurus coloradensis Knowlton ______X ____ ------~------Laurophyllum ripleyensis BerrY------~------X X ______------____ ----~------ Order MvuTALES Family MvuTACEAE Myrcia havanensis Berry ______X ------X ------X X Myrcia dubia Berry ______X Eugenia? anceps Berry ______.: ______X ------X --·------Myrtophyllum angustum (Velenovsky) Berry__ X X X X ------X Order UMBELLALES Family AuALIACEAE Aralia wellingtoniana minor Berry______X Aralia problematica Berry ______I X :x-1:::: ::::::::1:::: :::~ :::::::: ::~: :::~ :::: :::: 12 THE FLORA OF THE RIPLEY FORMATION Distributrioo of the Ripley flora-Continued Tennessee :Alabama Georgia ---- p., p., :::l 1=1 p., p., p., 1=1 ~ 0 0 l5 ~ 0 ~ ~ ~ s .!::: ~ 0 0 ·;:::: ::l ::l I:: gJ ~ ·a ::l ~ 0 0 o:s ..c:l o_e."' a:> z z"'' o. -e -e 0 0 .... 1=1 Ill 0 0 :g"' 0 0 0 0 o::l 0 Ill"' Ill p., 00~ :g ~ "' I:Q +>0 p., .:g~ p., ... +>p., I» gJ_o .:P.. I» p., ~P.. ~ t'' ,!4+> ra .. ~ t' ·~ • 1=1 ~Q =a:> ~ ;;~ 1=1~~ =a:>l=l..... ~"' or~ a:>l=l ::l "C)::l ~§ ...... ::l a:>::l >::l I:Q ·- ::l ~t' ~g z0 ::lo ... 0 I 0 <:(!0 fEO +>1=1 ao o:so 85 aS 1=10 "'0 :g .go oo !=10 >.0 0 a:> ~0 g~ ~0 ~8 a:> Ill ~ 0 0 CI.J l' Series GAMOPETALAE Order ERICALEs? Family ERICACEAE? Andromeda anceps Berry______X Order EBEN ALES Family SAPOTACEAE_ Chrysophyllum parvurn BerrY------,------X ------Bumelia prewilcoxiana Berry ______X --.------____ ---- ______Bumelia ripleyana Berry_· ______X ------____ ---- ______ Order GENTIAN ALES Family ASCLEPIADACEAE · Acerates cre:::il:e:::c~~~;~~~ ______X-~ _, ____ --- -~- ______,____ :______ Apocynophyllum giganteu~ Berry ______---- X __ ·__ ------'------______,___ ·_ ------Apocynophyllum perryensis Berry ______X ------______------Apocynophyllum sumterensis (Berry) Berry ___ - _ _ _ X --- _ - ___ ---- ______X ______Apocynophyllum ripleyensis Berry ______--- _ X ------_- - ~-- -- _- ---- ______- ____ _ Apocynophyllum alatum Berry ___ ---- _____ -- X ------·------______- ___ - ___ --- _- -- _ Position uncertain Calycites ripleyensis Berry ______----_-~---- X ------_ ------Phyll~tes hydrilloid~s J;3erry ______X ------1------Phyllites hydrochar1t01des Berry ______---- X ------Phyllites ripleyensis Berry ______---- X ------Carpolithus perryensis Berry ______----~_---- X ------:..------·- --- _ ------_ ---- Carpolithus ripleyensis Berry ____ .______------__ :.._ X ------Carpolithus carrollensis Berry ______---- X ------~------Halymenites major Lesquereux ______------.. ------X .. -- _ --- _ --- _ - ___ -.:..- _ The Ripley flora as at present known com in the future. Of the 135 species 86, or more prises 135 ·species, including the doubtfully or than 60 per cent, are new to science, and these ganic remains named H alymenites. Most of include the genera Dioscorites, Celtis, Cap these species are represented by well-preserved paris, Cedrela, Dillenites, and Chrysophyllum, and ample material and· are hence well char hitherto unknown in deposits older than the acterized ; fragmentai·y remains of still other Eocene. forms in the collection have been ignored. The total number of genera is 71, and these Undoubtedly more extensive collecting would are segregated into 40 families representing 28 greatly increase the nuniber of species, and orders. The supposed alga ( 11 alymenites) this is also to be expected should similar and the form referred to the genus Selaginella highly fossiliferous clay lenses be uncovered are of doubtful botanic affinity. The collec- DISTRIBUTION 13 tion contains very fragmentary remains of six .Myricaceae is third in size, with 11 different species of ferns. These are all small species, all of which ~elong to the genus and of more or less doubtful botanical iden 11/ yrica, which shows an extraordinary dif tity, but they are of importance for purposes ferentiation at this time. The next most of correlation, as nearly all are found in other abunda'ntly represented family is the Mor areas, especially in Greenland and Europe. aceae, with ten species, nine of which are. There are nine species of gymnosperms, but referred to Ficus and the tenth to a. striking several ·of these are of somewhat doubtful new form of Artooa1•pus. Non(3 of the species value in indicating age, arid traces of plants of Ficus are individually abundant in the of this affinity are extremely rare in western collections. Then follows the family Tennessee. The monocotyledons also number Celastraceae, with seven species. The family nine species, including a fan palm and a Apocynaceae has five species, all of which feather palm that is referred to the genus are referred to the genus Apocynophyllurn. Geonomitcs and is· of a type that is prominent The Euphorbiaceae, Rhamnaceae, and in the lower Eocene of the Raton Mesa region. Myrtaceae· have four species each, and the Several of the monocotyledons appear to rep J uglandaceae, Fagaceae, Ternstroemaceae,. resent aquatic or semiaquatic species. · Thus, and Sapotaceae have three species each. The there is a well-marked Potamogeton and the remaining families have only one or two mud~flat type Alismaphyl~'l.tm. In addition species each.. The largest single genera are the forms named Phyllites hyd'rilloides and 11/ yrioa, with 11 species, Ficus, with 9, and Phyllites hyd1·ooharitoides appear to represent Oelastrophryllum, with 6. If the type rep aquatics. The former shows great resem resented by Oelastrophyllum were considered blance to HydriUa, of the family fiydrochari with the forms that I refer to Ternstroemites, taceae, a genus supposed to be monotypic in which were formerly referred to Oelastro the existing flora and widely distributed phyllum, they would constitute the most throughout the Old World. The latter ap abundant type in both differentiation and in pears to represent an Upper Cretaceous type dividual abundance. of this same family, comparable in its habit Aniong the 105 species of dicotyledons of the. with the existing genus Vallisneria. Ripley flora 68 have entire margins. This is The dicotyledons of the Ripley flora num 64.7 per cent of th~ known forms. The ratio ber 105, ·of which 95 belong to the more of entire leaves to the whole flora of woody primitive or choripetalous series, and only dicotyledons has been shown by Bailey and 10 represent the more specialized Gamopet Sinnott 6 to have a fa1rly constant relation to alae. Among the Choripetalae the anemoph climatic conditions and altitude. According tc ilous forms, regarded as primitive by the figures compiled by these authors, over 70 some botanists and reduced by others, num per cent of the woody dicotyledons of sub ber 19. The largest alliance is the Rosales, tropical and tropical lowland floras haFe en with 14 species, all of which belong to the tire leaves. Their published percentagesrange Leguminosae. As .two of these species are from 71 per cent in the flora of Hongkong to definitely Teferred to the Mimosaceae, three to 86 per cent in the flora of the Malay States.· the Caesalpiniaceae, and three to the ~api Comparisong. of fossil with existing floras ate lionaceae, and as six can not be certainly always handicapped by the incompleteness of referred to any one family of Leguminosae, the fossil record, although, on the other hand: none of these families can be considered larger the vast majority of fossil forms are arbores than the others and all are less ·abundantly cent, so that the proble1n is simplified to that represented than some other families present extent. As the Ripley .flora is obviously a low in the Ripley : flora. The second laTg~st land coastal flora, the question of altitude is family is the La uraceae, with 12 species, eliminated. It would thus appear that the per but the recognition of minl.1te differences centage of entire leaves in the Ripley plants in the genera Oinnamomwm and .Lawrus accords well with the conclusions derived from tends to overemphasize their relative im 0 Bailey, I. ,V.,. and Sinnott.· E. W., Scien.ce, new ser., vol. portance in the Ripley flora .. ·The family 41, pp·. 832-833. 1915. 14 THE FLORA OF THE RIPLEY FORMATION the general facies of the flora that the climate From Cape Cod southward the shallow se~L was probably warm temperate and rather uni that covered the continental shelf overlapped form and the precipitation abundant. the margin of the Piedmont Pla,teau, and a The approximate outline of easter~ North broad arm reached up the Mississippi Valley as America during Ripley time is well shown in far as Illinois. The middle of the continent the accompanying sketch map (fig. 2). The was a region of shifting shore lines and shal North Atlantic, as recorded in the deposits low seas which con1pletely .sepa.rated eastern along the west coast of. Greenland, extended and western North America. Eastern North northward in the Davis Strait region between America was also separated from Central and FIGURE 2.-Mal> of North America showing the area of outcrop of the Ripley formation (f"Olid black) and the generalized relations of land (white areas) and sea (ruled areas) during Ripley time Greenland and Baffin Land. The extent of this South America by a profound expanse of arm of the sea is conjectural, as its sediments sea. have been discovered nowhere except in west Th'at the late Lower and earlier Upper Greenland. They were laid down in extremely Cretaceous floras of North America represent, shallow water, and presumably the gulf did in a broad way, a single botanic province is not extend northward much 'beyond the lati shown by the lack of dissimilarity between tude. of Disco Island, or else it was of a very the Albian, Cenomanian, and possibly Tur transitory nature, as there seems to have been onian floras of. the East and the West. There a free }UJnd path of migration between Green are naturally elements in ·each not found in land and the Atlantic coast of North America. the other, but these are to be attributed to the ANALYSIS OF OTHER .UPPER CRETACEOUS FLORAS 15 . imperfection qf the record or explained as types, common to either Greenland or Europe.. reflections of minor local facies. The time It would thus appear that the Ripley flora represented in the midcontinenta.l region by consists of some few forms existing in this the major portion of the Colorado and ~{on region. since Tuscaloosa time, with n1any new tana groups, however, was a time of independ elements that were added by immigration or ent evolution in. both the East and the 'iV est. evolution in the interval between the Tusca The 'iV est was broadly connected with the loosa and Ripley. gre'at land mass of Asia, and this fact might In so far as the present records permit a con .seem to furnish the setting for a greater re clusion,· the· Ripley flora may be regarded as semblance to the flora of Etirope than could the culmination of the earlier Cretaceous floras be sho,vn by the flora of eastern North Amer of eastern North America. Certain of its ica. This does not appear to have been the species are Ripley species only because of the case, ho:..vever, and .it would seem that there somewhat inconsistent usage of the forma was still a shorter and readily traversed path tional name Ripley by the Geological Survey. from the east across the Arctic and north At Thus Araucaria bladenensis and Ara~wa1ia lantic, and this paleogeographic conception jeffreyi, which are so highly characteristic would be equally true if we regarded the two of the Bl~ck Creek and Eutaw formations, :t·egions as receiving the bulk of their plant are recorded from the Ripley formation of population from a somewhat earlier Greenland Georgia. They have never been found in the center of radiation. true Ripley of Alabama, ~1:ississippi, or Ten It is interesting to see in this Ripley flora nessee, and as I have shown elsewhere in this a characteristic local facies, and this is es paper the Cusseta sand of the Ripley of pecially emphasized, as might have been ex Georgia is of the same age as the Coffee sand pected, by the dicotyledons. Nearly all the of the Eutaw of Tennessee. ferns and all the conifers have an outside dis The Ripley flora includes a considerable tribution, as befits their more ancient lineage number of forms that are cml.unon to the and earlier radiatio.n, but of the angiosperms Patoot beds of Greenland. Thus eight of the only 33 forms, or about 28 per cent, have been Patoot species are found in the Ripley, and found in other areas. This is still more .he general facies is more similar than the str:iking than appears from the bare s~'ate number of identical species would indicate. ment, for if Greenland is excluded as near the There are also eight species cm.mnon to the probable place of origin of the Upper Creta Vermejo flora of Colorado and New Mexico, ceous floras or at least the ancient home· of of which all but one are dicotyledon's and sev those of eastern North America, it seems that eral, such as the abundant Oissites pandu1•atus, only six of the angiosperms are common to have been found nowhere else. Tllis would dist'ant regions, such as Europe: These six serve to confirn1 l{nowlton's contention that are D1!JOphyllunL gracile, Ficus krausia.na, the V ermej o is older than Laramie, as the 11/ agnolia capellini, Euphorbiophyllwm anti only Ripley species recorded from the Laramie quum, OissUes crispus, and 11/yrtophyll·um an- iF Ll!y?ica to1·reyi, which also occurs in post gustum. . Laramie deposits in the West. Five species are The Ripley floi:a shows a marked relation found in beds in the 'iV est that are refm~red ship to that of the antecedent Eutaw formation to the Montana group, four in the ~{esaverde of this same region, to the Black Creek flora formation, two in the Fox Hills. and five of the Carolinas, and to the Magothy flora in the Fruitland formation. . of New Jersey and Mary land. As the known floras of the Eutaw and Black Creek are ANALYSIS OF OTHER UPPER CRETACEOUS FLORAS small and fragmentary this resembla.nce is GENERAL CONSIDERATIONS . all the more remarkable. Thus 16 of the Ripley species are common to the Eutaw, 19 Comparisons with European Upper Ci·e to the Black Creek, and 15 to the Magothy. ta~ceous floras show that five Ripley species are The older Tuscaloosa flora of the eastern Gulf recorded from beds on that contin~nt classed area has but six species that continue on into as Cenomanian, two fron1 beds classed as the_Ripley, and these are all widely distributed Turonian, three from beds classed as Em- 16 THE FLORA OF T'HE Rlf'LEY FORMATION . scherian, and six from beds that are either restricted to other regions befQre the dawn of. Santonian or Campanian. The Ripley flora the Eocene. Eocene prenuncial characters are contains noticeable elements common to that seen in the Ripley Geonomites, Dryophyllum, of the sands of Aix-la-Chapelle (Aachen), Celtis, Artocarpus, Cappa1·is; Dalbergia, Germany. How many -it is not possible to. Gleditsiophyllum, Oedrela, Dillenites, Tern .determine, for large collections made around stroemites, N ectandra, Oinnam,mnum, species Aix-la-Chapelle by Debey were never studied, retained in the form genus Laur:us, llfyrcia, and only the ferns have received adequate Chrysophylliwm, Bumelia, and Apocynophyl treatment. The plants from the plastic clays lu1n. Many of these are coastal types of Baume (St. V aast), France, described by that might be expected to persist in an Coemans, and found also at La Louviere, only slightly changed environment, but Belgium, are probably of the same age. There the same is true as regards the habitat has been considerable discussion and differ of most of the earlier Upper Cretaceous ence of opinion regarding the age of the sands floras of the Atlantic Coastal Plain, so of Aix-la-Chapelle. These sands contain a that these differences in the Ripley flora sparse shallow-water marine fauna with as· compared with antecedent floras may Exogyra, ·Trigonia, Eriphyla, lp,oceramus, truly be regarded as of chronologie value. and lenses of clay containing fossil plan,ts. Detailed cmnparisons with the extensive Wil The sands are overlain by a highly fossil cox flora (Eocene) show many striking dif iferous glauconitic sand-the Acti1wcarnax ferences. In the 'Vilcox the old Mesozoic quadratus zone of the Campanian. This re coniferophyte elements are all gone, and nu lationship would seem to indicate that the un merous genera of angiosperms found in the derlying plant-bearing beds represent the Ripley, commonly considered as temperate littoral and continental deposits of the San forest types, such as Salix, Fagus, Celtis., Liri tonian or at most can not be older than the odendron, Platantts, Acer, Oornophyllum, lower Emscherian Coniacian substage. De and Andromeda, were appaJ:'ently lacking spite this, Haug refers these sands· to the along the Eocene coast . of the Mississippi Turonian. They are siliceous and cross Gulf, although presep.t in the Rocky Moun bedded and in part represent· coastal dune tain and Great Plains region at that time. sands, so that they might conceivably be con This would seem to indicate some climatic siderably older than the overlying strictly; difference between Ripley and Wilcox condi marine se~iments. The question is important, tions,. and in that event. the Ripley would be be.cause I regard the Ripley as of about the more distinctly temperate than the Wilcox. same age, but perhaps it can not be settled in The most extensive, best-preserved, and the present state of our know ledge. I am dis earliest studied Upper Cretaceous ·floras, both posed, however, to regard them as representing in this country and abroad, have been: those the Santonian substage, or upper Emscherian. in the continental deposits in1mediately pr~. The known ·Ripley flora contains no forms ceding the transgression ·of the Upper Cre that are common to the abundant Wilcox r;r taceous sea, or in the initial deposits of that lower Eocene flora of the Mississippi embay sea-such, for example, as the Dakota sand meilt region, the only forms doubtfully re stone of Kansas and Nebraska, the Rarita~.l corded from the Eocene in other regions being clays of .New Jersey, the Perucer beds of Geonomites schimperi and the wholly worth Bohemia, or the Niederschoena deposits of ~ess lialymenites major. Despite this lack of Saxony. In all the sections the plant evjdence identical species, which might be expected, the rapidly diminishes in both variety and qual Ripley flora does contain a number of. types ity from botton1. to top.· A knowledge of the that bet::a-me. differentiated subsequently and Ripley flora not only serves to offset this con are characteristic elements of the lower Eo dition and furnish a basis for more precisely cene flora in southeastern North America. evaluating the passing of Cretaceous time These are all angiosperms, the genera and but it introduces us to many of the smaller species of the pteridophytes and some of the and more delicate forms., such as almost in coniferophytes having become extinct and variably fail of preservation in coarse de some genera of the coniferophytes having been posits. ANALYSIS OF O'l'HER UPPER CRETACEOUS FLORAS 17 Examples of the relative paucity of the Cretaceous time the floras, as might be pre later Cretaceous floras of the Atlantic Coastal dicted a priori, were essentially of Jurassic Plain are furnished by noting that the Tusca type. Only gradually did they become typi loosa, Raritan, and J\1agothy floras compris~ cally Cretaceous, and those elements that give about 150 speci~s each, whereas in my report them a Cretaceous facies lingered in gradu on the Upper Cretaceous floras of the world,7 ally diminishing numbers to the end of Upper I listed but nine· species from the Ripley Cretaceous time. Floras as early as the Al formation in Georgia, two from beds of that bian show a variety of dicotyledons, including age in Alabama, and three from Tennessee. many still existing genera, but the numerical . A foreign parallel may be taken from the representation. of dicotyledons in the rocks Bohemian area. I-Iere the Perucer beds ( Ceno tends to be illusory, and a large number of manian) have furnished 184 species, the Turo the Cretaceous dicotyledons are more or less nian 4 7, and the Chlomeker ( Emscherian) synthetic types, if one is justified in drawing but 32. such conclusions from the evidence of foliage It m~1y be noted that of the genera. identified alone. The lower Eocene introduces us to in the Ripley flora. 35~ or 51 per cent, are ex floras that are of a decidedly different facies tinct. . ~1ost of these, which are angiosper and that distinctly show a great Cenozoic mous, are form genera and therefore not as modernization, _despite the difficulty of reflect precise as might be wished, although ·obviously ing this change in the current nomenclature some, such as 1'e'l"nstroemites, represent the of paleobotany. stock from which the Tertiary and existing Moreover, when it comes to details the genera were differentiated, and it is note change is profound. If the Ripley flora is worthy that 16 of the genera have not been compared with the vVilcox flora of the same recognized in post-Cretaceous deposits. The general region the two are ~een to be almost following are the ext_inct genera of the Ripley totally unlike, though living under simila.r flora: environmental· conditions. Both are lowland Halymenites. Caesalpinites. coastal floras, both are warm temperate in Raphaelia. Gleditsiophyllum. type, and yet they have no common species. A Dryopterites. Leguminosites. tyro would immediately recognize the one as Monheimia. Euphorbiophyllum. Cretaceous and the other as Tertiary.. The Taeniopteris. . Manihotites. Cretaceous cycadophytes and conifers have Protophyllocladus. Cissites. Widdrii.1gtonites. Grewiopsis. disappeared in the interval between the two; Cunninghami tes. Dillenites. the flowering plants are distinctly better dif Moriconia. Ternstroemites. ferentiated and more modern. Geinitzia. Laurophyllum. That these Eocene floras as well as the Upper Alismaphyllum. 1\fyrtophyllum. Doryanthites. Cornophyllum. Cretaceous dicotyledons .had ancestors is no Dioscorites. Acerates. more pertinent than the fact that the Eocene Geonomites. Apocynophyllum. mammals had ancestors. In fact, the com Dryophyllum. Calycites. parison is not inept, because the evolution of Men isperm i tes. Carpolithus. the flowering plants was,. I believe, the 1najor Acaciophyllites. Phyllites. Mimosites. factor that made possible the evolution of the mammals. That students in general have underesti· The other U ppe]: Cretaceous floras in various mated the contrast between Upper Cretaceous parts of the world that may profitably be com and Eocene floras is, I think, obvious. It is pared with the Ripley flora are shown on the the fashion to picture the incoming of the accompanying sketch map (fig. 3). Some of Angiosperms in ·Upper Cretaceous time· as a these are of great importance, such as those sudden and overwhelming event, which in of Montana age in the western interio.r region Cenomanian time transformed the J\1esozoic of North America, or of especial interest be floras of ferns, cycadophytes, and conifers into cause of their geographic position, such as that an essentially Cenozoic flora of broad-leafed of the Patoot beds of Greenland. Others are modern forests. At the beginning of Lower worthy of comment because they are of ap 7 Maryland Geol. Survey, Upper Cretaceous, 1916. proximately the same age-as, for example, the 18 THE FLORA OF THE RIPLE·Y FORMATIO.N I plants of the Izumi series of Japan,' of south- .are unknown in the Ripley, as well as the eastern France, of the Alpine Gosau beds, and twigs and scales of Araucaria, which, al o~ the Aix-la-Chapelle sands-even though as though recorded from the Ripley, come fron1 to the~e flora.s our information is still very deposits of the same. age as the Eutaw in incomplete. other regions, a~ I have already explained. EU'fAW FLORA These are preserved in the Eutaw because of their resistance to maceration, but they should The first comparison that naturally sug be present also in the Ripley if they had not gests itself is· a comparison with the flora· of become extinct.. Similarly among the dicoty the Eutaw forn1ation, whose deposition imme ledons .the prevailing forms found in the diately preceded that of the Ripley formation Eutaw are coriaceous types such as Androm in this area. The Eutaw flora is still incom eda, Oinnamomurn, Fimts, jJfagnolia, M yrcia, pletely known. The Eutaw sediments in so jJfanihotites, and Scobalites. far as they might preserve land plants suc In the following list of the known Eutaw cessfully are either marine clays or littoral, species those marked * are also present. in the prevailingly sandy deposits. The clays con Ripley flora. Of the 46 Eutaw species 16 con tain only decay-resisting forms such as Ara'n- tinue into the Ripley, but the following genera FIGURE 3.-Sketchi map of the world showing the location M fossil floras that have been compared with the Ripley flora. 1, Ripley; 2, Black Creek and Magorl1y; 3, Vermejo ;. 4. Montana group·; 5, Vancouver; 6, Patoot; 7, Izumi series; 8, Por tugal ; 9, Aix-la-Chapelle; 10, Marseille, Fuveau, and Toulon; 11, Sub-Hercynian and Westph3.Iian; 12, Chlomeker beds of Bohemia and Silesia; 13, Gosau beds caria twigs and coriaceous dicotyledons, and of the Eutaw have not been recognized in the the sands contain similar forms that succes3- R.ipley: Androvettia, Brachyphyllurn, Oepha fully withstood the triturating action of the lotaxospermum, Oupressinoxylon, Dewalquea, troubled waters, as no clay lenses with any Diospyros, P aliurus, Pterospermites, Sequoia, extensive representation of the plants that and Tumion. gr:ew along the Eutaw shores have yet been Andromecla cretacea Lesquereux:? (Ga.) discovered. Andromeda novaecaesareae Hollick. (Tenn.) The known Eutaw· flora, which has been Andromeda parlatorii Heer. (Ala.) Andromeda wardiana Lesquereux:. (Ga., Tenn.) . collected chiefly from the Chattahoochee Val Androvettia elegans Berry. (Ga.) ley, in Georgia ; near Havana, in Hale Aralia eutawensis Berry. (Ga.) County, Ala.; and near Beacon, in Decatur *Araucaria bladenensis Berry. (Ga.) County,· and at Coffee Bluff, in Tennessee, *Araucaria jeffreyi Berry. (Ga.) consists of but 46 species. It includes no ferns, Bauhinia alabamensis Berry. (Ala.) Bauhinia cretacea Newberry. (Ga.) 34 angiosperms, andll gymnosperms. Among Brachyphyllum macrocarpum formosum Berry. the gymnosperms are the genera A ndro (Ga.) v_ettia, Brachyphyllum, 06phalotaxospermum, Cephalotaxospermum carolinianum Berry. Oupressinoxylon, Sequ_.oia, and Tu.mion, which (Ala.) ANALYSIS OF OTHER UPPER CRETACEOUS FLORAS 19 Cinn:unomum heerii Lesquereux'! (Ga., Tenn.) Pachystima? cretacea Berry. *Cinnamomum newberryi Berry. (Ga.) Laurus atanensis Berry. Cupressinoxylon sp. Berry. (Tenn.) Apocynoph~•llum sumterensis (Berry) Berry. Dewalquea smithii Berry. (Tenn.) Diospyros primaeva Heer. (Tenn.) .These also should probably be regarded 9.S *Doryanthites cretacea Berry. (Ala.) forms which in all probability were present *Myrtophyllum augusta Velenovsky. (Ga.) in the Eutaw flora but were not preserved or *Ficus crassipes ( Heer) Heer. (Ga., Tenn.) have . not been discovered. *Ficus krausiana Heer. (Ga., Tenn.) Ficus ovatifolia Berry. (Ga., Tenn.) MAGOTHY FLORA *Geonomites schimperi Lesquereux. (Tenn.) *Halymenites major Lesquereux. (Tenn.) The Magothy flora, which is found from Juglans arctica Heer. (Ga.) Marthas Vineyard to Maryland, is much Laurophyllum elegans Hollick. (Tenn.) larger than either the Black Creek or the Lam·us plutonia Heer. (Ala.) Magnolia boulayana Lesquereux. (Ala., Ga.) Eutaw flora. The beds in which it occurs *Magnolia capellinii Heer. (Ga.) are more remote from the Ripley area geo *Malapoenna horrellensis Berry. (Ga., Tenn.) graphically and aiso in part definitely older. *Manihotites georgiana Berry. (Ga., Tenn.) It is of especial interest because in the north *Menispermites variabilis Berry. (Ga.) ern Atlantic Coastal Plain the · fossil plants *Myrcia havariensis Berry. (Ala., Tenn.) Paliurus upatoiensis Berry. (Ga.) are found in large. part in the Magothy for Phragmites pratti Berry. (Ga., Tenn.) mation, which is transgressed by the Matawan Phyllites asplenioides Berry. formation-a marine glauconitic deposit that Pterospermites carolinensis Berry. (Tenn.) has yielded only two or three fossil plants Sabalites carolinensis Bei-ry. (Tenn.) whereas in the Carolinas and the Gulf region Salix eutawensis Berry. (Ga., Tenn.) Salix flexuosa Newberry. (Ga.) alternation of conditions· or near-shore phases Salix lesquereuxii Berry. (Ga.) of the marine formations have furnished :t Sequoia ambigua Heer. (Ala.) considerable number of fossil plants of the Sequoia reichenbachi (Geinitz) Heer. (Ala., same genern1 facies throughout a 1nuch greater Ga.) thickness of sediments. *Ternstroemites ripleyensis Berry. (Tenn.) Tum ion carolinianurn Berry? (Ga.) The Magothy flora, as at present known,· *Zizyphus laurifolius Berry· (Ga.) consists of 195 species, many of which repre Ripley plants that have been found also at sent surviving Cenomanian types. A large pre-Eutaw horizons and are therefore legiti number of these survive from the underlying Raritan formation. In 1914 I made the state n1a~ely considered as probable unknown mem 8 bers of the Eutaw flora are ment that there were no known Cenomanian floras in North America, unless the Raritan Widdringtonites reichii Ettingshausen (Heer). flora of the East and the Washita flora of the Moriconia cyclotoxon pebey and Ettings- est represented that stage of the European hausen. vV Moriconia americana Berry. section, and that the post-Raritan and pre Myrica brittoniana Berry. Ripley floras of the East and the Dakota sand Leguminosites canavalioides Berry. stone flora of the West were of Turonian age. Nothing that has come to light since. this state BLACK CREEK FLORA m.ent was written has rcast any doubt upon its The second flora of interest in this connec c.orrectness. On the contrary, additional in tion is the Black Creek flora of the Ca.rolinas. formation serves to strengthen the opi~ion, and This is of approximately the same age as that I would even be inclined to include the upper of the Eutaw and contributes the following . Ra1~itan flora in the Turonian category, al elements to the Ripley flora that have not though the New Jersey Geological Survey still yet been discovered in _the Eutaw: maps the Raritan formation as Lower· Cre taceous. Fifteen species of the Magothy flora Protophyllocladus lobatus Berry. Mor:conia americana Berry. are found in the Ripley. None of these, how Cunningh:unites elegans (Corda) Endlicher. ever, are especially abundant or characteristic Myrica brittoniana Berry. 8 Berry, E. W., The Upper Cretaceous and Eocene floras of Ficus celtifolius Berry. South Carolina and Georgia: 1!. S. Geol. Survey Prof. Paper Celastrophyllum carolinensis Berry. 84, p. 71, 1914. 20 THE FLORA OF THE RIPLEY FORMATION of the Magothy with the exception of certain Legundnosites, Platanus, Rhmnnus, and Zizy long-lived forms like Widdringtonites reichii, pitus. The Monocotyledonae number 5, and Ounn.inghamites elegans, Fio~ts krausiana, one of these-a Pota;mogetor~r-is close to a Ficus crassipes, M agnoUa capellini, and Oin Ripley species. namomum newberryi. Elements of the Mago There are 20 species of ferns in the Patoot thy· that apparently became extinct before flora, and among them species of Raphaelia Ripley time are the numerous species of An and Asplenium, are cominon to the . Ripley. dr·mneda, Gleichenia, H ymenaea, Laurus, The conifers number 18 species and include Liriodendron, Magnolia, 11/yrsine, Pinus, Po che genera Oyparissidium, Dammara, Oun dozamites, Quercus, Salix, Sassafras, Sequoia, ninghamites, Geinitzia, Glyptostrobus, lno Sterculia, lVilliamsonia, and Zizyphus. lepis, 11! oriconia, Pinus, Sequoia, T axites, and W iddringtonites. Five different species of PATOOT FLORA Sequoia are recognized. Five. of the conifer Interest in the Cretaceous Arctic floras has ·ophyte genera occur in the R.ipley, and the always been very great and ·is enhanced by the species of Ounningharnites, M oriconia, and theoretic possibility of their representing the W iddrin,qtonites fr.om the two regions are original center of radiation of the Upper Cre identical. taceous angiosperms. The stratigraphy of There is little in the Patoot flo1·a to mark western Greenland has rested almost entirely its occurrence 2,400 miles farther a\-ray fr01n on the paleobotanic work of Heer, who segre the Equator than the Ripley flora. A slight gated the somewhat similar lithologic serie~ difference might p,erhaps be predicated fro111 . exposed there into three · units-the !Come, the absence of palms or Artocarpus and the Atane, and Patoot beds, entirely upon the evi abundance of conifers in the Patoot flora sup dence of their contained floras. Subsequent plemented by the presence in the Ripley flora exploration has shown that the limits of these of such genera 'as Dioscorites, Oapparis, Dal units· are exceedingly doubtful. and that. al be'rgia, Ternstroemites, · Bumelia, Oedrela, though the main bodies of these three floras Dillenites, and Ohrysophyllum, which are sup.: are distinct, the boundaries of the units are posed to ·indicate a warm habitat and which largely arbitrary. There rcan be no doubt that are not represented at Patoot. the ICome flora is Lower Cretaceous or that There has been some difference of opinion the Atane flora belongs in the early half of regarding the age of the Patoot flora. The as the Upper Cretaceous. The Patoot flora, sociated invertebrates were classed as Senonian which occupies a stratigraphic position be· by De Loriol, and VVhite and Schuchert in tween the Atane flora and that of the overly~ cline to the belief that what Heer called the ing Tertiary-Heer's Arctic Miocene-is the Patoot beds are transitional both stra.tigraphi-· one that is of 1nost interest in the present con~ cally and paleontologically from Cretaceous nection. to early Tertiary. The few invertebrates of The Patoot beds of western Greenland ap either· the Atane or the Patoot beds are of pear to be purely a paleobotanic unit,· if they slight significance, as Stanton has stated, and really merit the term unit, because lithologi therefore the conclusion that these beds are cally they can not be differentiated from the somewhat near the age of the Pierre and Fox older Atane beds or the younger Tertiary de .I-Iills of our "\Vestern States is wid1out any posits. The Patoot flora is 'a relatively large satisfactory evidence, although possibly true. one, comprising about 125 species. Of this The flora, as perhaps overdifferentia.ted by number 77 are dicotyledons, of which 7 are· Heer, appears to have a reasonable degree ·of refer!·ed to Quercus and o.nly 2 to 11/yrica, al hon1ogeneity and distinctness of facies, ·al though some of the supposed oaks are cer though it contains a la.rge number ( 34) of tainly Myricas. Species of lJtf yric.a, Oinna forms that are comma~ to the Atane beds. nimnum, and La~trus are identical with Ripley The flora contains 20 Dakota species, 22 forms, and, in addition, the following di-· Raritan species, 19 Magothy. ~pecies, 8 Tusca cotyledonous genera are represented in both loosa species, and 4 Black Creek species. There floras: Acer, Acerates; Aralia, Carpolithus, are 11 species common to the Perucer beds of Oelastrophyllum,· Oissites, Ficus, Juglans, Bo~wmia and Moravia, 6 to Nicderschoena, 7 to ANALYSIS OF OTHER UPPER CRETACEOUS FLORAS 21 the European Turonian, 3 to Aix-la-Chapelle and that would increase·the number of identi (Aachen) and 3 to the vV estphalian Cretaceous cal forms. Probably some of the small forms (Campanian). Four have been identified in and leguminous leaflets represented in the the ~1ont.ian of Europe. Ripley flora are due to the fact that the Ripley The large number of Atane species present, was a coastal flora, but aside from this environ as well. as the numerous Dakota, Raritan, and mental factor my experience indicates that in ~1agothy species, preclude considering the flora general lithified sediments are less productive as young as, for example, the Laramie. It is in well-preserved remains or small forms than singular, if the Patoot flora is younger than fine clays like those of the Ripley in northern the Emscherian, that it should have so much Tennessee. more in common with the Cenomapian and VANCOUVER ISLAND Turonian floras than with the extensive Lo\ver (Campanian) and Upper (Maestrichtian) From the Upper Cretaceous of Vancouver Aturian floras so ·extensively developed 'in Island CN anaimo, Port MeNeill, Bayne~ · northern Germany. On the other hand, the Sound) and the probable extension of these four Montian species are not without signifi beds on Protection and Newcastle islands the cance. It is possible that several horizons of late Sir 'iVilliam Dawson recorded a consid erable flora of about 80 species. These in- t~1e Upper Cretaceous a.re represented in the Patoot, although this seems doubtful. .eluded 10 ferns, 2 .cycads, 6 coniferophytes, An· interesting feature of similarity between 3 monocotyledons, two of which are fan palms, the Patoot and Ripley floras is the abundance and 58 dicotyledons. Two ginkgoes are re of dicotyledonous leaves with· toothed mar corded, as well as six species of Ficus~ and six species are referred to gins-a feature which, according to the statis Quercus. This flora appears to have little in common tics compiled by Bailey and Sinnott, is indica with the Ripley flora and seems to include tive of nontropical conditions. some discordant elements, but this may be VERMEJO FLORA due to the lack of precision in the identifica tion and illustration of the material. The Vermejo flora of southeastern Co.lorado COLORADO GROUP and northeastern New Mexico as recently n~onographed by ICnow lton ° comprises 108 The flora of the Colorado group of the forms-1 fungus~ 5 algae, 14 ferns, 9 conifers, western interior is, unfortunately, too imper 5 monocotyledons, 60 dicotyledons, and 14 fectly known to afford profitable comparisons. forms that were .not specifically identified. . It It comprises 24 species described by l{no'w lton contains 17 species of Fi(Jus, of which only one from the Frontier formation of Wyoming, 2 is common to the Ripley. Other prominent species of fungi from the Graneros formatioD elements are 8 species of V ibur1];um~ and 5 of Nebraska, and 18 species described by Daw species of Salix. · Ferns are more numerous son from the Mill Creek formation of British and varied than in the Ripley. Genera found Columbia, whose age is in doubt· and may b~ in the Vermejo but not represented in the somewhat older. The Ripley flora contains Ripley include Abietites, Acrostichwm, Ame no species common to these florules, nor do lanohier, Ane1nia, Brachyphyllum, Oanna, these in fact show any especial similarities to Oolutea, Oredneria, Diospyros, Fraxinus, the pre-Ripley floras of the Atlantic Coastal Gleichenia, H edera, 0 smunda, P alaeoaster, Plain, so that one is forced to conclude that Phaseolites, Polystichu1n, Pteris, Pterosper the similarities between the Dakota flora of 'lnites; Que'l·c'lts, Seq'uoia, Viburnum, and Wood the West and the Raritan, Tuscaloosa, and ~oa1·dia. The two floras have eight identical ~1agothy floras of the East were not con species, and doubtless there were many small tinued through Colorado time. or delicate forms in Vermejo time that failed of preservation or detection in the relatively MONTANA GROUP coarse material in which the plant fossils occur With the florules of the Montana group the Ripley has more in common. Thus the Mesa ° Knowlton, F'. H., TJ., S. Grol. Survey Prof. Paper 101, pp. ~42-:.:83, 1917. verde formation contains 5 species common to 22 THE FLORA OF THE RIPLEY FORMATION the _Ripley, the Vermejo. formation contains so named from the nwunta.in range of that 8, the Judith River formation 2, the Trini name along the west coast, and beds of similar dad ·sandstone 1, the Fruitland formation 5, age (Upper Cretaceous) are widespread the IGrtland formation 1, and the Fox Hills 2. throughout Japan,: in pla,ces overlying the The so-called Laramie flora as it stands in Ryoseki series. They contain a fauna of which the literature is composite, and until the re Trigonia pocillijor1nis Yokoyama is the most cent publication of the revision of these floras characteristic species. Fossil plants ha~e been by Knowlton/0 stu,dents _have had difficulty in reported from a nun1ber of localities and are making correct evaluations. The revised list said to include Sequ.oia, Populus, Sa;lix, Q1tB1' of plants from the Laramie formation of the cus, Fagus, Platanus, Oinnamo1nJU111, Arundo, Denver Basin as given by l(nowlton 11 shows and other genera. More detailed information only a single species, 11! yrica torreyi, common regarding this exceedingly interesting flora is to the Ripley. This species is a v~riable one greatly to be desired. that n1ay really be composite, and in any event it does not detract from the obvious weight PORTUGAL of the evidence, which shows that the Ripley Similarly in Portugal a considerable flora and Laramie floras are extremely dissimilar. from beds of Senonian (probably Emscherian) These differences are so great and of such age has been noticed by DeLima, but no de a character that they indicate not regional tails beyond tentative generic identifications variation but difference in age. Thus the are available. Among these the genera com Laramie flora contains a variety of forms mon to the Ripley are Oinnanwmun1, Dam prenuncial of Eocen3 floras, such as Anenvia, mara, Sabalites, La;urMs, Afagnoli(J), Afyrica, Anona, Aristolochia, Artocarpus, Oeanotl~us, and ZizyphMs. 'Additional information re Oerci8, and Do'fnbeyopsis. The many species garding this flora is also grea.tly to be desired. of Ficus, Juglans, Laurus, Phyllites, Rhiun as the earlier Portuguese floras of the Albian nus, Salix, and Zizyphus are all different from and Cenomanian and still older Cretaceous those of the Ripley. I conclude that the Rip show such a striking parallelism \vith those of. ley flora is of pre-Laramie age and that the our Atlantic Coastal Plain. Laramie of the West is .represented in the At lantic Coastal Plain by the interval between GERMANY the Cretaceous and the Eocene. This conclu sion, it seems to me, is as secure as it is possible A full account of what is known of the to make interpretations of the past, and it is of flora of the sands of Aix-la-Chapelle and especial interest because those who make cor near-by regions across the border of Rhenish relation tables from the literature rather than Prussia was published in 1916.12 The age from the study of forn1er life have frequently of this flora is somewhat uncertain, but it is correlated the Ripley with the Lara-mie. of note in that it contains six species common to the Ripley-three ferns, two conifers, and JAPAN a dicotyledon--and three of these are also Comparisons between the Ripley flora and common to the Patoot beds of western Green those of other continents naturally are of less land. value than comparisons with floras of the same A considerable number of fossil plants have land mass, and this is especially true of com been described from the Harz region (sub parisons with the floras of higher stages of the I-Iercynian) and Westphalia in Germany, but Cretaceous, because most of the foreign work their resemblances to those of the Ripley are has been either not done at a-ll or else is very on the whole not_ especially close. Fossi~ old and sadly in need of revision. For the first pla-nts are not uncommon in the Emscherian of these reasons it is not possible to compare of the Saxon-Bohemian border region (Chlo the Ripley flora with the fossil plants of the meker beds). In addition to more or less Izumi series of Japan. The Izumi sandstones, worthless remains, 32 species, mostly well characterized, have been described, and these 1° Knowlton,. F. H., The Laramie flora of the DP.nver Basin : U. S. Geof. Survey Prof. Paper 130, 1922. 12 Berry, E. W., Maryland Geo1. Survey, Upper Cr_etaceous, 11 Idem, pp. 9'5-96. - pp. 266-271, 1916. . ANALYSIS OF OTHER UPPER CRETACEOUS FLORAS 23 include two, Oissites crispus and My?·tophil more closely allied to the Eutaw and Black l'un?J angust'l.tm, that are common to the Ripley. Creek floras of North Alnerica, but it is of importance in preventing the assignment of a GOSA U FLORA too young age to the Ripley. "'\Vith the European floras referred to the 'I'he Gosau beds of the Tyrol· and eastern l\1ontian stage the Ripley flora has nothing in Alps contain fossil plants at a number of lo common. calities, as at St. Wolfgang, "'\Viener-Neustadt. Brandenberg, and Gri.inbach. These are of BEARING ON CORRELATION interest because the Gosau beds, in part at Long-distance correlations are obviously un least, are the equivalent of the Ripley. The certain, particularly if they are interconti published list of species is small and shows nental, and although experience has shown that no especial affinities with Ripley !orms, but the march of evolution of the Upper Creta much undescribed material from Griinbach is ceous floras was remarkably parallel in North reported by !Crasser to include a number of Ripley genera. America and Europe, the absence of reliable data at present from the European region ren FRA:NCE ders precise comparisons out of the question. Plant-bearing deposits at a number of lo After a careful consideration of the evidence calities in southeastern France-namely, the available I am of the opinion that the Ripley Turonian near Marseille (Var), the Enl flora is of Emscherian age, although it might scherian of Beausset, near Toulon (V ar), and be as young as the Campanian substage of the the Aturian lignites of Fuveau (Bouches-du Aturian. I am strongly of the opinion that it Rhone)-are of interest in the present con~ec is not as young as the Maestrichtian or upper tion, though here again the floras have hot substage of the Aturian. As regards our been adequately described. The chief simi western Cretaceous section the Ripley flora in larities to the Ripley are to be found in the dicates contemporaneity with the l\1ontana Turonian, which contains two identical species, group. More precise correlation is not possible Eupho1·bioprhyllum antiquum and W idd1ing at the present time. t01~ites reichii, and a number that are similar The appended table shows the members of for example, in Arauca;ria, Artoca:rpus, Dry the Ripley flora that have an outside distribu ophyl7Jwln, llfenispermites, and 11/yrica. This' tion, with ·the occurrence and range of each. resemblance does not indicate identity in age, It serves as a basis and graphic summary of as the known Turonian floras of Europe are the preceding discussion. 24 THE FLORA OF THE RIPLEY FORMATION 'Ripley plants found in other formations 0.. ~ 0.. ;;; 0.. a> .,;, a, .a .... .!!.l a> ~ ce ....~·~ ~::~ 6~::~ .2 ~ § .2 j:. o § .2 ~::~ .~ 'g § § ~ .2 ~=: ::: gj ~ § ._ ~::~ 1:1 .2 1:1 ~ .;9 .ce § .~ .§ 0~ ~.~ ~~ .9 ~'CJ ~ VJQ) .9~ '0 ~.9 ~.s a:>:z:; o.9 ~~ <1).9 ~·en a ~ ·;:::: §§ 1 . ~ ~ {1 ~ i ~ ~ § ~ ~ ~ ~ ~ j ~ ~ ~ ~ ~ ~ ~ ~ ~ § ·~~ ~ ~ ·~ ~ ~!I ~ i ·~ i ] i ~ ,.:g.S ~ :l.S ~ . ::_ .s C\l 5 .!:::; 2 0 rn ~.2 ~ ~ ta ~ g a> ::; t:l §o ~I=Q,.,.E-t~:s: f:OA;tl ~~~~"'~"'<...:!Jl.i ~OE-t~w. t__l_}_j-~-:~-1-:~-~.-~-:-e~-~n-fo-J-~-~-l-~-~d-~e-~--~-~~-~-~-~ ~~~ ~~~ =~~ ~=h~ ~~~ ~=~ ~~~ ~~~ ~~~ ~~- ~J=~ =~= =~~ j:~~ ~~~ ~~~ ~~~~-~- 1 1 Protophyllocladus lobatus ____ X X ------·--~'------~--- Widdringtonites reichiL_:. ______X X X'--- ___ ? ______---'--- X ______Moriconia cyclotoxon_.:. ______X "'-- X I ___ --·- ______, ______X 1 Moriconia americana ______X X ___ ---'------__ .. ------~--- 1 Geinitzia formosa ______------X X X!------x· --- X I X X~------X Araucaria bladenensis _____ X X X ------'------'------Araucaria jeffreyi______X X ______---~--- ___ ,_:._---~--- ___ I ___ ---~--- ______-·- __ _ Cunni.nghamites elegans______X X _ _ _ X _____ - ______X __ - _____ - X _ _ _ X __ _ Dammara acicularis ______X ___ ---,--- ______· Doryanthites cretacea_ __ _ _ X X X ------Geonomites schimperi_ ____ X ______------? ------======{¥~~~~:Myrica torreyi_js~~~~f;t~pi~~======______-~- ______X X X -~- X ___ X X ______Myrica brittoniana ______X X ___ ------Salix gardneri______X ______Dryophyllum gracile ______;, ______:.. ______- ______X X Ficus krausiana ______X X X X ______--- ___ ------X X ------ Ficus crassipes ______X X X X _:..._ --- X X ------7 ------Ficus celtifolius ______X ______------_..:_ ------Ficus leei ______X X ______X ______Liriodendron laramiense ___ ·--- ______--- -~------X ------~:~~~!~~·ffe:1~~~i;bili~=== ~ -~- -~- -~- ======-~~ ======·======~======-~- ======~ ;t~~~~~~~~~;~~~~~~ :x: ::: ~x~ -~~ ~~~ ~~~ :x: _·_ ~~~ -~- .-r~ ~-: -~- .:~ -~- ~~~--~E-x~~:--~- 8f:1lf::t~~~~f~~:~~~-:-~::: ::: :?: ::: ::: ::: ::: ::: ::: ::: :::1::: ::: ::= :~: ::: ::: ::: =:: ::: :::.? ::: Zizypl1us laurifolius ______X ______------'------Ternstroemites ripleyensis_ _ X ______-- ___ -· _-- --,... ------Cinnamomum newberryL __ X X X X X ? ___ X ------·------Malapoenna horrellensis ___ X X ______------Laurus atanensis ______X X ___ --- ______? ------·------t:~~~: ~~i:~J~~~f;~======::= === ::= ::= -x.- ::= -~- -x- -~- ::: ::= ::: ::: ::: ::=1::: Laurophyllum ripleyensis ____ ·____ -~- ______--- --L ___ X --- ___ , ___ ------: Myrcia havanensis ______X ____· ______: __ ------·------Myrtophyllum angustum __ X X ______--- ______------X ------Apocynophyllum sumter- H:l~~~~it~~~~j;r____ ~===== -x_- -~-======I======-x_------x_- -x_- -x_------x_- -x_- ======LYCOPC:>DIALES, FILICALES 25 SYSTEMATIC DESCRIPTIONS Phylum PTERIDOPHYTA Phylum LEPIDOPHYTA ~ Order FILICALES Order LYCOPODIALES ~ Genus ASPLENIUM Linne Asplenium calopteris (Debey and. Ettingshausen) Family SELAG INELLACEAE ~ Beer Genus· SELAGINELLA Beauvois~ Plate I; figures 9, 10 Selaginella laciniata . Lesquereux . Ren·izia oalopteris Debey and Ettingshausen, K. Sel.a.ginella laoiniata Lesquereux, U. S. Geol. and Akad. Wiss. Wien Denkschr., Band 17, p. 216. Geog. Survey Terr. Bull., vol. 1, p. 378, 1875; (36), pl. 5, figs. 13-17, 1859. U. S. Geol. and Geog. Survey Terr. Ann. Rept. 1a Bayer, Edwin, Einige neue Pflanzen. der Perucer Kreide Rock, Carroll County, Tenn. schichten in Bohmr.n, p·. 45, text fig. 14, pl. 2,. fig. 2, 1899'. Fl'ic, .Anton, and Baye1·, Edwin, Naturwiss. ·Landesdurch 14 Knowlton. F. H .. Flora of the Montana formntio~: U. S. forschung Bohmen .Archiv. Band 11, p. 70, text fig. 6, 1901. Geol. Survey Bull. 163. p. 19, pl. 3, figs. 1, 2, 1900. FILICALES 27 Genus MONHEIMIA Debey and Etti~gshausen I in the region of the sinus., but normally a vejn . detaches itself from the midvein at a very Monheimia aquisgranensis Debey and Ettingshausen t I d' f d' t d · acu e ang e, ascen 1ng or some IS ance an Plate I, figures 1, 2 after giving rise to two branches turning sharply outward and proceeding to the mar ll.'lonheintia aq1tisgmnensis Debey and Ettingshausen, gin i,n a course parallel with that of the two Die urweltlichen acrobryen Kr~idegebirges Aachen : K. Akad. Wiss. Wien Denkschr., Band branches. The lamina is not strictly linear 17, p. 211 (31), pl. 4, figs. 3-10, 1859. but expands slightly dist'ad. The following is the original descriptiOn· o f Occurrence: MeN airy sand member of the Ripley formation, Cooper pit, near Hollow this species : Rock, Carroll County, Tenn. M. fronde bipinnatipartita vel pinnatisecta, laciniis nrrecto-patentibus vel arrectis, alternis, remotis, de Genus TAENIOPTERIS Brongniart currentibus, lineari oblongis, anguste frondosis, 1-3 milliill. circitm· longis, 1lh-5 millim. latis, apice acumi Taeniopteris sp. nntis, integerrimis; rhachibus costisque Ionge flexu osis, tenuibus, glabris, subteretibus; soris in summi Plate I, figures 3~ 4 tatibus laciniarum copiosis, ma~imam fronclis an- gustae laminam tegentibus, a costa mecliana remoti- In the collection from the Cooper pit, near oribus subhaemisphaericis. Hollow Rock, there are several fragments of Although it is true ·that Debey and Ettings.,. the lamina of a linear frond, pinna, or pinnule, hausen describe sori which they take to indi- ' which are readily recognizable and yet too in- cate a polypodiaceous affinity, and no traces of complete for adequate diagnosis. They are of sori are seen in the Tennessee material, they a coriaceous texture and range from 1 to 1.6 were inclined to see and idealize fructifica centimeters in width, with a fairly prominent tions in poorly. preserved material from Aix midvein, · which is not more prmninent on la-Chapelle and in other material where such either the upper or lower surfaces. The lat features were nonexistent, as for example in erals are stout, largely immersed in the leaf the gymnosperm lJf oriconia, so tha·t in the ab substance; they dive.rge from the lateral mar sence. of actual specimens one may well doubt gins of the midvein at wide angles, forking the accuracy of both descriptions and figures. once almost immediately, and proceed in par· The Tennessee material is scanty, and I have allel courses to the margins. Weathering of no opinion to offer regarding. its botanic affin the thick substance of the lamina often gives ity. It would be without intere11t were it not the appearance of. a crenate margin, but I for the significan~e attached to the appear have satisfied myself that the margins are ance on both sides of the Atlantic at about entire. · the same ti1ne of a variety of identical. plant The question of the generic reference is fossils, of whieh this is a peculiar example, one on which opinions will differ. 1\::now.t .. tha.t have not been detected in the anrecedent ton 15 has referred similar but somewhat rocks of either rr.gion. The present species larger fragments from the Montana group. of also shows simila,rities to the figures of an Wyoming to Asplenium, but it se~ms better to other from Aix-la-Chapelle which Debey and ·refer the present material to Taenioptem~ Ettingshausen call Zonopteris goepperti, but which is after all a form genus without any as that forn1 is based exclusively on what is especial significance regarding botanic rela said to be fertile material and is described as tionship. having a crenate margin, the American mate By Brongniart's definition 16 the present rial is considered to be that of illonheimia. fragments belong in Taeniopteris. Heer 17 de It may well be questioned if a number of spe scribed a Taeniopteris deperdita from the cific and generic distinctions in the classic pa Upper Cretaceous of Greenland, but this was. per by these authors are not more finely drawn based on an uncertain fragment which is at thaii the facts warrant. . . In the Tennessee material refei·red to M on 111 Knpwlton, F. H., Flora of the Montana formation.: U. S. Geol. Survey Bull. 163, p. 20, pl. 3, fig. 11, 1900. ~uAn~ia the venation is well preserved and con ~6 Brongnlart. A. T., Prodrome d'une hlstoiL·e. des v~getaux s:ists of relatively stout veins that are. invari fossiles, P'· 61, 1828. · . ·" 11 Heer,. Os,vald, Flora fossilis a.rctica, vol. 7, p. 8, pl. '48, ably forked. There mn,y be a single fork, a::: fig. 14, 1883. . THE FLORA OF THE RIPLEY FORMATION least specifically different from the Tennessee eral lobes. Axial vascular strand very stout below, ::;pecimens. These are not dissimilar to llf a becoming very thin and finally disappearing by re from the Upper Cretaceous of peated branching apically. When the leaves are Tattia cTetacea lobate, subordinate opposite vascular strands form the . 18 Bohemia, described by· Velenovsky. There ax.is of the lobes, and these also are. generally but is also s.ome resemblance to an entire-margined not invariably lost before reaching the tips of the Nilsonia, as, for example, Nilsonia bohemica, lobes by giving off innumerable secondary branche.s. from the Upper C~etaceous of Bohemia~ de Margins in.·all specimens are rather remotely undu 9 late-crenate, and the tips are all rounded. Sec Rcribed by Velenovsky/ but this resemblance ondaries numerous and thin, diverging from the main is purely superficial, as Nilsonia has simple axis of the phylloclad or the axis of the lobes at very \cP.ins and the lamina at the top of the ra =:his. acute angles, curving outward, some simple but many Occurrence : l\icN airy sand member of the dichotom.ously forked, and a few several times Ripley formation, Perry place, Henry County, forked. Lobes when present are separated by cuneate, narrowly rounded sinuses which terminate some dis ~tnd Cooper pit, near Hollow Rock, Carroll ·tance from the main axis. The largest specimen, Couuty, Tenn. which is still incomplete Loth at the apex and at the base, measures 8 centimeters in length and 5 centi Family POLYPODIACEAE meters from tip to tip of the lower lobes, the entire Genus DRYOPTERITES Berry upper portion measuring about 1.5 centimeters in width. Dryopterites stephensoni Berry These remains are superficially like fern fronds, especially in specimens which are compound, and D1·yopterites stephensoni Berry, U. S. Geol. Surve_y 'vere it not for the presence in the Cretaceous of Prof. Paper 84, p. 103, pl. 17, figs. 1, 2, 1914. other Phyllocladus-like remains with a demonstrated This species is fully described in the publica gymnospermous structure (for example, A nd1·ovettia) tion cited aboye, and it has not been discovered their reference to this genus would seem hazardous. outside of the type locality .. The entire specimens are strikingly like some of the Occurrence: Cusseta sand member of Ripley forms of Protophyllocladus subintegritolius ( Les quereux) Berry, of the Raritan and Magothy forma formation, cut on Central of Georgia Railway tions, or like Protophyllocladus polymorph~ts ( Les llf2 miles northeast' of Byron, Houston County, quereux) Berry, from higher western American hori Ga. zons, and even the compound specimens have an un lobed apical portion of comparable length which is Phylum CONIFEROPHYTA also similar in appearance to the two species just Order CONIFERALES mentioned. The compound forms are superficially like Thinnfelclia rhomboidalis Ettingshausen,20 the Genus PROTOPHYLLOCLADUS Berry type of the genus Thinnfeldia, whose systematic posi tion has been the occasion of so much controversy and Protophyllocladus lobatus Berry which has been variously regarded as a fern, as u cycad, and as a conifer. The present species shows 7'hinnfeldi.a n. sp. Berry, Johns Hopkins Univ. Clrc.,· . important differences, however, aside from being new ser., No. 7, 1907, p. 81. much younger, and it is confidently believed to be Protophyllocladtts lobatus Berry, Torrey Bot. Club unrelated to the ·various older Mesozoic species of Bull., vol. 38, 1911, p. 403; U. S. Geol. Survey 1'hinnfeldia which have been described . . P.rof. Paper 84, p·. 17. pl. 2, figs. 9-13, 1914. It may also be. compared with various forms from The present record is based on a few frag the Upper Cretaceous of Dalmatia discussed at great ments in the Ripley formation of Tennessee. length by Kerner,21 who refers them to the genns The species was originally described from Pachyptcris, which he regards as cycadaceous in nature. rather complete material from the Middendorf The present species is believed to be closest to arkose member of the Black Creek formation Protophyllocladus subintegrifolius, a species which in South Carolina. Following is a quotation is abundant in. the Atane beds of Greenland, the of the original, description : Dakota sandstone of Kansas and Nebraska, the Raritan formation of ·New Jersey, and the Magothy Leaves (phyllocluds) of large size, lanceolate or formation from Marthas Vineyard to New Jersey, and oval in general outline, either entire with crenate which commonly assumes a sublobate form. This is margins, rounded apex, and narrowly cuneate base or especially shown .in unreported collections made by · compound through the development of opposite lat- the writer in the ·l\Iagothy formation of New Jersey. 18 Yelenovsky, Josef, Farne der bohmischen Kreideforma :lO Ettingshausen, C. von,. K.-k. geol. Reichsanstalt Abh., tion, p. 9, pl. 1, fig. 13, 1888. vol. 1. pt. 3, No. 3, p. 2, pl. 1, figs. 4-7, 1852. 111 Velenovsky, Josef, Gymnospermen der bohmischen KL·ei 21 Kerner, F. von, K.-k. geol. Reichsanstalt Jahrb., vol. 45, deformation, P- 11, pl. 2, figs. 25-28, 1885. p. 3!), 1896. CONIFERALES 29 No new light is shed upon the habit or sys This species, which is not rare in the Ripley tematic position of the genus by the meager formation of western Tennessee, is probably amount of material in the collection from the n~ost common conifer of the Raritan for T'ennessee. mation. It v;ras described originally by Et Occurrence: MeN airy sand me1nbei· of Rip tingshausen from the Cenomanian of Nieder ley formation, Perry place, Henry County, schoena, in Saxony, as a specit:s of F1·enelites. Tenn. 'Vhen Heer discovered it in the Greenland Genus WIDDRINGTONITES Endlicher 1naterial, where it ha,s been collected from both the Atane and the Patoot beds, he tran~ Widdringtonites· reichii (Ettingshausen) Heer 22 ferr·ed it to the present genus. It has subse Ji'·rcnel'ites 1·eioh·i·i Ettingshausen, Kreicleflora von quently been reported from the Cenomanian Nieclerschoena, p. 12 (246), pl. :.t, figs. 10a-10c, of Bohemia and ]\;fora via, from the l\1agothy 1867. lVidcl1'•ingtonites ·mich'i-i Heer, Flora fossilis arctica, .formation at numerous localities, and from the vol. 6, Abt. 2, p. G1, pl. 28, fig. 5, 188~. sot~thern New England islands. It has .also Berry, Maryland Geol. Survey, Upper Cretaceous, been reported from the Tuscaloosa formation p. 793, pl. 55, fig. 1, 1916. ' ·of Alabama, where it is abundant at a number l\1eclium-sized branches with more or less of localities, and is apparently represented in crowded, slender, elongated fastigiate twigs, the Dakota sandstone by the remains which bearing reduced ovate-subulate leaves, spirally Lesquereux referred to Glyptost1·obus. arranged. Both mici·osporangiate and mega Occurrence: l\1cN airy sand Inember of" sporangiate cones have been found. The cone::; Ripley formaaon, Cooper pit, near 1-Iollow are small oval bodies 5 to 12 millimeters long Rock, a quarter of a mile west of Buena by 3 to 7 millimeters in cha1neter, usually pcor Vista, Carroll County, Tenn. ly preserved, said by Ettingshausen to. be axil lary in position but evidently often terminal, Genus MORICONIA Debey and Ettingshausen as evinced by some of the Raritan material and [K. Akad. ·wiss. Wien Denkschr., Band i7, p. 239, by so1ne of the better-preserved cones from the 1859] Cenomanian of Bohemia and l\1or.a.via. The Leafy twigs, apparently deciduous in habit, latter material clearly shows that the cones bifacial, phylloclad-like, made up of dec.ussatc consisted of four scales. This would ally it 4-ranked leaves, those on the flat faces round either with the subgenus lVidd1'ington.ia of the ed, those folded at the margins pointed, the genus Oallit1'i8 Ventenat, to which Eichler in whole resulting in a st:r:ictly geometric pat his treatment of the living species in Engler tern. No traces of fructifications known. and Prantl ( 1887) refers Endlicher's genus, or The type of the genus was described in 1859 to the subgenus E'U,oallit1'is Brongniart, which by Debey and Ettingshausen in their account also is characterized by four cone scales. Eu of the ferns fron1 the sands of Aix-la-Cha oallitris has a single living species of northern pelle. They. consideted illm'ioonia a fern, "as Africa, and lV idd1·ingtonia has three or four did Heer the first specimens collected from species of southern Africa and ~1adagascar. Greenland, and there is no doubt· that speci l'he propriety of Eichler's classification may mens in which the outlines of the leaves a're well be questioned, and in any event paleobot indistinct are very fernlike in appearancr.. anists must necessarily prefer the older segre Saporta pointed out the coniferophyte nature gation of F1·enelal and lViddrvngt'on.ia and of this genus in 1868, comparing it 'vith Libo their respective fonn genera. oed1"tt8 and Thuiopsis, 23 and most subsequeil't There seems to be little doubt that the pres P.uthors, including Schenk, Newberry, and ent species should be. refernicl to 111 idd1·in,r;· Lange, hav'e referred the genus to the Cupres tonia, as Velenovsky and !Crasser have done. sinaceae. Two species have been recognized, but as the term TVidd?·i~ngton~£tes ~s equally in although the differences betwe·en them are clicative of its true affinity, little is to be gained slight. I-Ieretofore these two forms have by making the proposed change. never been found associated, and this fact ::::The extensive synonymy of this species has J?een given 2a Saporta, Gaston de. Soc. geol. France 1\Iem., 2d ser., vol. In several recent publications and need not be repeated .. s. p. 301, 1868. 100913-25---3 30 THE FLORA OF THE RIPLEY FORMATION . largely influenced their separation. Now, with seven cycles of leaves. They decrease however, both forms are known to occur at somewhat in length upward and then increase the same horizon in the Ripley formation of considerably, so that distad they have a length Tennessee-at a single locality in the MeN airy of 17 millimeters and show about 12 cycles of sand member. There seems to be little doubt leaves. A detached fragment of this form that these two expressions of 11! oriconia. habit ·from the locality in Henry County, also fig represent the same or closely related species ured, has similar narrow lateral twigs, but and that the differences are due to the relative these are about 3 centimeters in length with age or position of the twigs. They certainly 17 or 18 cycles of le~ves. In all the material do not represent either chronologie or geo the leaf covering of the main axis, though pres graphic variations. At the same time the ent in ·juvenile forms, is less markedly de two forms are strikingly i1nlike in appearance, veloped than in 1.l/. americana and tends to be and after mature consideration it seems best come obliterated both by age and by fossiliza to emphasize this difference in appearance by tion, so that· it is rarely preserved in detail as different names, as it seems probable that the in that species, although its character is wen· larg.er mutants are progressing in the direc shown in some of the specimens frmn the tion of the genus Androvettia:. Raritan formation of New Jersey figured by Newberry. · Moriconia cyclotoxon Debey and Ettingshausen The species was described originally from . Plate III, figures 1, 2 specimens obtained in the sands of Aix-la Chapelle (Aachen), and its superficial resem Moriconia, cyclotoxon Debey and Ettingshausen, K. blance. to a fern, especially when poorly pre Akad. Wiss. Wien Denkschr., Band 17, p. 59, pl. 7, figs. 23-27, 1859. served, led Debey to insist always on its fern Hee·r, Flora fossils arctica, vol. 6, Abt. 2, p. 49, nature. Similarly the first specimens dis pl. 33, figs. 1-9b, 1882 ; idem, vol. 7, p. 11, pl. covered by Heer in the Greenland Upper Cre 53, fig. 10, 1883. taceous were described as a species of Pecop Lange, Deutsche geol. Gesell. Zeitschr., Band 42, teris. There can be no doubt, however, of its . p. 665, pl. ~m. ·fig. 4, 1890. Schenk, Palaeophytology, p. 318, [ig. 220, 1890. gymnospermous nature. Newberry, U. S. Geol. Survey Mon. 26, p. 55, Subsequently. the species was recorded from pl. 10, figs. 11-22, 1896. the Saxon ·Cretaceous (Turonian~) , from the · Hollick, New York Acad. Sci. Annals, vol. 11, Atane beds ( Cenon1anian) and Patoot beds p .. 57, pl. 3, fi~. 10; p. 418, pl. 37, fig. 8, 1898; (Turonian or Emscherian) of Greenland, from U. S. Geol. Survey Mon. 50; p. 46, pl. 3, figs. ·16, 17, 1907. . ' the Raritan formation (Cenomanian) of New Berry, New Jersey Geol. Survey Bull. 3, p. 8G, Jersey and Staten Island, and from the Mag pl. 8, figs. 3-6, 1911. othy formation (Turonian) of Block Island. Pecopteris kudlisetensis Heer, Flora fossilis arctica, It has not been' found elsewhere in our At vol. 3, Abt. 2, p. 97, pl. 26, fig. 18, 187-1. lantic Coastal Plain until its recent discovery Twigs of somewhat varying size and pro in the Ripley formation of western Tennessee, portions, consisting of a thick central axis, in beds which I regard as probably Einscherian bearing two strictly opposite rows of more or in age. There has been considerable difference less reduced or extended, abruptly pointed of opinion regarding the equivalents in the lateral twigs, the whole covered with closely \Yestern United States and in Europe of the appressed .decussate leaves, those on the faces · Coastal Plain Upper Cretaceous formations, short and evenly rounded, those terminating invertebrate paleontologists leaning to the at the margins slightly longer and pointed. view that they were younger than the paleo Substance not at all thickened, as in the genus botanic evidence seemed to ·admit. Brachyphyllum, but thin, although not as thin From its bearing upon the place of origin as in the allied species 11! oriconia americana. and Upper Cretaceous dispersal of floras the The material from the Ripley formation of determination of the oldest horizon at which Tennessee shows long, slender forms like the 11! oriconia cyclotowon has been found is a mat larger one figured, in which the narrow lateral ter of considerable interest. This, it seems to twigs are only about 2 millimeters in width me, is in the Atane beds of western Greenland. 1nd 10 to 12 mill~meters in length, proximad These beds were fully discussed by Heer, who CONIFERALES. 31 considered them the equivalent of the Ceno Occurrence: l\1cNairy sand member of Rip manian of Europe. The few poorly preserved ley formation, Perry place, I-Ienry County, and invertebrates associated with the plants arc Cooper pit, near Hollow Rock, Carroll County, stated by "'\Vhite and Schuchert 24 to indicate 'I'enn. strongly an equivalence with the. Pierre and. Moriconia americana Berry Fox I:Iills of the western United States and \ . 'v.ith the Senonian of Europe. This is not the Plate Iii, figures 3, 4 place for an analysis of this statement, but it jJ1oriconia u1nericana Berry, Torrey Bot. Club Bull., would seem that a decided paleozoologicopinion vol. 37, pp. 20, 186, 1910; U. S. Geol. Survey might well await the discovery of a represen Prof. Paper 84, p. 26, pl. 7, figs. 1-4, 1914; tative fauna, and -that meanwhile the evidence Maryland Geol. Survey, Upper Cretaceous, p. of a flora of 182 fossil plants, 36 of which are 802, pl. 56, fig. 1, 1916. .Morioonia cyclotoxon Berry, New York Bot. Garden common to undisputed Cenomanian localities Bull., vol. 3, p. 65, pl. 43, fig. 4; pl. 48, figs. ·1-4, ir~o Europe (Bohemia; ~f··Tavia, and Saxony), 1903 [not Debey and Ettingshausen] ; Torrey might be considered as having· some bearing Bot. Club Bull., vol. 31, p. 70, 1904, vol. 33, pp. on the question .. Although dogmatism is to be 165-167, 1906. deprecated, I can not co.nceive that the general Leafy twigs, consisting of a main axis and relations of these floras permit their being con short, reduced, lateral branches of decussate sidered younger than Turonian, and in any leaves. Along the main axis on each flat face event the present species of JJ[ m'ico1-da fully of the branch these leaves are relatively and confirms the general~y accepted notion of a closely appressed, with' a narrow base and a northern origin and southward migration of broad semicircular apex. The corresponding these eai'lier Upper Cretaceous floras. · lateral pairs of leaves are thin and pointed As yet no traces of illO'riconia have been and transversely compressed. In the axis of found in the abundant plant beds of the Upper each of these marginal leaves is a reduced Cretaceous of the western interior; of North . Lranch flattened in the same plane as the main America nor in the upper Cretaceous of eastern branch, so that the. whole arrangement is Asia, recently differentiated on Sakhalin Island strictly opposite and distichous. These re by ICryshtofovich. The specimens from the duced lateral branches have leaves of the same Atane beds figured by I-Ieer in 1882 show con character and ·arrangement ~s those of the siderable variation, particularly the very great main branch. The bifacial leaves are, how extension of some of the· lateral twigs shown ever, somewhat smaller and blunter, and the in his Plate XXXIII, Figure 4, where they marginal leaves are broader and less acute. I are twice as long as the balance of the laterals They become rapidly smaller distad, it usu on the same branch. His Figure 2· on the same ally l'e(~uiring not more than five or six pairs plate shows a· form of branching not observed to complete the blunt lateral reduced twigs. in the rest of the JJ[ oriconia. material and re The main vascular axis is stout, but that of motely suggests a comparison with the sup,. the .lateral twigs is thinner and much less posed· staminate aments of Androvettia. It is pron1inent. The leaves fail to show any veins. true that as drawn by I-Ieer this small lateral The texture was apparently coriaceous but ob cluster has the same form and arrangement of viously thin in most of the specimens. The parts as the nor1i1al JJ! miconia twig, hut I-Ieer Tennessee material shows evidence of a slight was not~rious.ly careless about detail and in thickening along the keel toward the apex. of illustrating poorly preserved material often the lateral leaves and along the truncated supplied the deficiency of the specimen, and in margin of the bifacia.l leaves.· The leaves are this case he might well have assumed the pres concrescent, and the lowermost lateral of each ence of the central axis and regularly decussate subm:dinate twig has disappeared on the distal leaves, whereas this branch may really have side, its position being occupied by the long been like those of A nd1·ovettia. that have been and· narrow ma.rginal of the main axis. The considered to represent staminate sporophy lls. great majority of specimens have extremely reduced lateral twigs, but in the specimen .of· 2' White, David, and Schuchert, Charles, Cretaceous series · this species · from the Magothy formation of of the west coast of Greenland: Geol. Soc. America Bull., vol. !), p. 366, 1898. New Jersey th~se laterals are extended and 32 THE FLORA OF T'HE RIPLEY FORMATION · reach a length. of 3 centi1neters and a width Berry, New York Bot. Garden Bull., vol. 3, p. G7, of 9 millimeters and consist of eight or nine 1903 ; Torrey Bot. Club Bull., vol. 31, p. 68, pl. pairs of leaves. 4, figs. 2, 3, 1904; New Jersey Geol. Survey Bull. 3, p. 97, 1911; Maryland Geol. Survey, M oriconia americana is known only from Upper Cretaceous, p. 801, pl. 54, fig. 6, 1916; the Atlantic Coastal Plain, having been found U. S. Geol. Survey Prof. Paper 112, p. 61, 1919; previously in the Magothy fonnatibn of New Geinitzia sp. Newberry, New York Lye. Nat. Hist. Jersey, Delaware, and Maryland; the Black Proc., vol. 2, p. 10, 1873. Creek formation of North Carolina; and the ~equoia r.eichenbachi Lange (in part), Deutsche geol. Gesell. Zeitschr., Band 42, p. 770, 1890. Middendorf arkose n1ember of the Black Stanton and Knowlton, Geol. Soc. America Bull., Creek formation of South Carolina. The oc YOl. 8, p. 137, 1897. currence in the Ripley formation of Ten Sequoia gracillima Newberry, U. S. Geol. Survey Mon. nessee greatly extends its range, both geo 26, pl. 9, figs. 1-3, 1896 [cones only]. graphic and geologic, and indicates its prob- . Berry, New York Bot. Garden Bull., vol. 3, pl. 48, figs. 21, 22, 1903 ; Torrey Bot. Club Bull., vol. able occurrence in the intervening region be 31, p. 69, pl. 2, 1904; Am. Geologist, vol. 34, pl. tween South Carolina and Tennessee, as well 15, 1904; Torrey Bot. Club Bull., vol. 32, p. 44, as in the intern1ediate Eutaw formation, but 1905; idem, vol. 33, p. 165, 1906; New Jersey it has not yet been found at this horizon Ol' State GeologiRt Ann. Rept. for 1905, p. 139, in these regions. · jJf oriconia, especially jJJ. 1906. mnericana, is strikingly like the genus Geinitzia graoillima Jeffrey, Bot. Gaz., vol. 51, pp. 21-27, pl. 8, 1911. .._ 25 Geinitzia formosa Heer · ent in appearance, and it is believed tluJ,t ma terial of this species in that condition of pres Gein'itz'ia to1·mosa Heer, Zur Kreideflora von Quedlin ervation was the basis for the Raritan forms burg: Schweizer. paleont. Gesell. Neue Denk- · schr., Band 24, p. 6. pl. 1, fig. 9; pl. 2, 1871. that were identified as jJficroza1nia gibba 26 Newberry, U. S. Geol. Survey Mon. 26, p. ~1, pl. H, (Reuss) Corda by Newberry. A single cone fig. 9, 1896. is contained in the M'agothy collections made Hollick, New York Acad. Sci. Trans., vol. 16, p. along the Chesapeake & Delaware Canal in 129, pl. 12, figs. 1, 2, 1897. Delaware. Knowlton, U. S. Geol. Survey Bull. 163, p ..28, pl. '5, figs. 1, 2, 1900; U. S. Geol. Survey Prof. The foliage, which resembles somewhat that Paper 98, p. 333, pl. 85, fig. 3, 1916; U. S. Geol. of Sequoia reichenbachi (Geinitz) Heer, as Survey Prof. Paper 101, p. 251, .pl. 31, figs. 1-3, well as that of Ounningharnites squ.amosus 1918. Heer, shows rather thick t'~igs with slender ~5 This species is said to be represented by OarJJOUthes curved needle le'aves interspersed with small hcm.locin1ts of Schlotheim. See also Otto·. Ernst von, Addit amente zur Flora des Quadergebirges in der Gegend urn 26 U. S. Geol. Survey Mon. 26, p. 45, pl. 12, figs. 6, 7, Dresden und Dippoldiswalde, pt. 1, pl. 5, figs. 1-·3, 5, 6, 1896. See also Berry, E. W., New Jersey Geol. Survey Bull. 1852. 3, P'· 78, 1911. CONIFERALES 33 scalelike leaves. This foliage is unquestion Genus CUNNINGHAMITES Presl ably distinguished with difficulty from that of Cunninghamites elegans (Corda) Endlicher Seq1wia 'reichenbachi ( Geinitz) Heer, and it 0 unning hamites elegans (Corda) Endlicher, Synopsis is very likely that some of the records of. the coniferarum, p. 270, 1847. latter, which are so extensive, both geographi Newberry, U. S. Geol. Survey Mon. 26, p. 48, pl. 5, cally and geologically, include occurrences of :figs. 1-7, 1896. Hollick, New York Bot. Garden Bull., vol. 2, p. Geinitzia. An extensive review of the liter 402, pl. 41, fig. 11, 1902. ature 'vould be necessary to d~termine this Berry, New York Bot. Garden Bull., vol. 3, p. 64, point, and even this might prove fruitless. 1903 ; Torrey Bot. Club Bull., vol. 31, p. 70, pl. In general the twigs of the Sequoia 'are more 3, figs. 7, 8, 1904; New Jersey Geol. Survey Ann. slender with stouter leaves all of one kind, Rept. for 1905, p. 138, 1906. Knowlton, U. S. Geol. Survey Bull. 257, p. 135, pl. whereas in Geinitzia the b~igs are stouter and 15, fig. 1, 1905. the leaves are longer and more falcate and are Hollick, U. S. Geol. Survey Mon. 50, p. 41, pl. 3, interspersed with small scalelike leaves. Ex fig. 1, 1906. cept for the last feature these criteria 'are of Berry, Torrey Bot. Club Bull., . vol. 37, pp. 186, uncertain value and lead to much confusion. 505, pl. 20, figs. 1-4, 1910; U. S. Geol. Survey Prof. Paper 84, pp. 24, 106, 1914; Prof. Paper The cones are, of course, very readily dis 112, 1919. tinguishable, but except for their. presence at This widespread species has not been found · the type locality and their prodigious abun in the Ripley except in the Cusseta sand mem dance in the Magothy formation around Rari ber near Byron, Ga., which, as I have shown, tan Bay they 'are not represented with the is the chronologie equivalent of the Coffee.sand foliar remains, so that all these identifications member of the Eutaw formation. It is essen are hence open to more or less doubt. tially a, somewhat older species, but as it has For the foregoing reasons any extended been recorded in the Judith River beds of comment on the distribution of Geinitzia for Montana there is no known reason why it nwsa is not warranted. The type material should not range up into the true Ripley in came from the Senonian of Saxony, and there the Atlantic Coastal Plain. is no do~1bt that it is represented in the Mag Occurrence: Ripley formation, Cusseta sand othy formation of the Atlantic Coastal Plain. member, near Byron, Houston County, Ga.. Purely foliar records include the sands of Aix la-Chapelle, the Montana group in 'Vyoming, Genus ARAUCARIA Jussieu the Vermejo formation in Colorado, and the Araucaria bladenensis Berry Fruitland formation in New Mexico. It · would· seem singular that there .aJ:e no Green Arauca1·ia bladenensis Be1·r1J, Torrey Bot. Club Bull., land records of Geinitzia, because Greenland vol. 35, p. 255, pls. 12-14, figs. 1-3, 1908; ( ?) vol. 38, p. 405, 1911; U. S. Geol. Survey Prof. ahnost invariably constitutes one of the three Paper 112, p. 61, pls. H, 10, 1919. regions furnishing records of Cretaceous forms This species is fully described in the publi common to North America and Europe. This cations cited above. It is exceedingly common is readily explained, however, by the probable in and characteristic of the Black Creek· assumption that some of the numerous foliar formation in North Carolina. In South Caro · remains from Greenland which I-Ieer referred lina it is found in the extension of these beds. to one or another of his numerous species of It is present in the lower part of the Eutaw Sequoia actually represent Geinitzia. In the formation and in later Cretaceous deposits in absence of specim@ns it is not worth while to western Georgia and along Chattahoochee base any conclusions on figures of these species. River. Careful search has failed to discover Occurrence: J\1cN airy sand member of Rip this species in the fossiliferous plant beds of ley :formation, Cooper pit, 11ear I-Iollow Rock, western Alabama of Tuscaloosa age, but it Carroll County, Tenn. is present in great abundance at the. very base 34 . THE FLORA OF THE RIPLE:Y FORMATION of the Eutaw deposits in Hale County. Re Upper Cretaceous forms described under other cent collections have shown it to be present i:p. specific names-for. example, Dammara bo're ·the Magothy formation in Maryland. Singu alis Hem·-so · that specific characters other larly enough, it is not found among the abun than those of size, shape, and stratigraphic dant display of Ripley plants in western Ten position are not available. nessee, a fact which suggests that it does nut Remains of these organs were described range above the Eutaw and that the Cusseta and figured by Hitchcock in his account of sand of Georgia should be correlated with the the organic. remains found at Gay Head, Eutaw rather than with the Ripley formation. Marthas Vineyard, as long ago as 1841. He Occurrence: Eutaw formation (basal beds), did not name them but remarked: " It seems 2 miles south of Havana, Hale County, Ala. ; to me very obvious that these remains must Chimney Bluff, Chattahoochee County, Ga. be the seed vessels ·of some coniferous plants." Ripley formation · (Cusseta sand member) , In 1882 Heer found similar forms in the ma Buena Vista, Marion County, Ga. terial from the west coast of Greenland and named and described three very similar forms Araucaria jeffreyi Berry and definitely recognized their relation to Arauca1·ia )effreyi Berry, Torrey Bot. Club Bull., vol. Dammara. Subsequently they have been re 35, p. 258, pl. 16, 1908; U. S. Geol. Survey corded from the European Cenomanian by Prof. Paper 84, p. 105, 1914. Velenovsky, J(rasser, and Beyer, from the Raritan formation of New Jersey by New These cone scales. were· completely described berry, and . from Long Island and. Staten in the papers cited and need not be discussed Island by Hollick. They are abundant in the here. middle part of the Raritan at Woodbridge, Occurrence : Eutaw formation,. Chimney N. J., and occur in the upper part of the Bluff, Chattahoochee County, Ga. Ripley for Raritan (at the same horizon as the beds at mation (Cusseta sand member) , near Byron, South Amboy, N. J.) i1nmediately across Houston County, Ga. Arthur Kill on Staten Island. Genus DAMMARA Lamarck I have found these scales in the 1\fatawan formation of Maryland, in the Black Creek Dammara acicularis Knowlton formation of North Carolina, and in the Tusca Plate XXII, figure 8. loosa formation of Alabama. They .have not been detected in the Cretaceous deposits of Dammara acicularis Knowlton, . U. S. Geol. Survey Bull. 257, p. 134, pl. 15, figs. 2-5, 1905 ; Canada · South Carolina or Georgia. · Geol. Survey Summary Rept. for 1915, p. ·205. Similar remains have been considered by Heer, White, J(rasser, and others as represent This species, which was described from ing the fruits of the Eucalyptus, but it seems specimens collected from · the Judith River obvious that their relations are definitely with formation of Montana and which has also the araucarian conifers. been recorded from the Belly River formation Occurrence: McNairy sand member of Rip of western Canada, is represented by a single ley formation, a quarter of a mile west of specimen in the Ripley. It was described by Buena Vista, Carroll County, Tenn. J(nowlton as follows: · Cone scales broad and very thick at apex, nar Phylum ANGIOSPERMOPHYTA rowed below into a broad, thick basal portion ; apex Class MONOCOTYLEDONAE broadly rounded, provided in the center with a long, • slender, acute, apparently depressed awn; body of Order N AIADALES scale with several (8 to 10) strong apparently con centric ribs [resin canals], which are pressed close Family NAIADACEAE together and pass down the narrowed basal portion Genus POTAM:OGETON Linne as thin str:ae. Potamogeton ripleyensis Berry, n. sp. This species is abundant in the Judith River beds, and Knowlton regards the awn Plate III, figure 5; Plate XXIII, figures 1-3 as a distinctive character. Similar awns are, Leaves small, oblanceoiate, with a pointed however, occasionally present· on the earlier apex and a narrowed apetiolate base. Texture NAIADALES 35 thin. Length 5.1 centimeters; maximum tude-for example, Potamogeton perfoliatus width, above the middle, 1.2 centimeters. Ve Linne extends over more than 20 degrees of nation aerodrome; veins numerous, thin, paral latitude in America, frmn Newfoundland and lel, rarely forking, immersed. C~'OSS veinlets British Columbia to Florida and California, present. and also occurs in Europe and Asia. It is an This new species, superficially suggestive of interesting fact that all the wide-ranging Podoz.amites ~n· some cycade~n pinnule, is of species extend into both comparatively high a thin texture and lax· venation entirely con and low latitudes, whereas species of more forming to Potam,ogeton. Conclusively. identi restricted range, such as Potamogeton fiori fied Cretaceous remains of this genus are rare, danus Small and Pota1nogeton curtissii Mo the only other Cretaceous species kno\vn to me rang, of Florida, are commonly confined to being Potamogeton cretaceus Heer 27 and warmer regions. This condition may be taken Potanwgeton m/tddendorfensis Berry/8 both to indicate either that species confined ·to low of which are very similar to the Ripley species. latitudes in the existing flora, . or their im Both are sessile or subsessile. The former is mediate ancestors in more ancient floras, had more symmetrically lanceolate, somewhat originally a n1uch wider range than now, or h:.rger, and relatively broader, with coarser that the modern wide-ranging species have and much fewer veins. It comes from the ap greatly extended their range in recent times. proximately synchronous horizon of the Pa I incline to accept the latter supposition, al toot beds in western Greenland. The latter though it is well known that the majority of is shotter and relatively wider, with a broadly. aquatics, both animal and plant, are little in rounded tip, and comes from· the Middendorf fluenced even by rather wide differences in arkose member of the Black Creek formation, latitude.· · in South Carolina, a slightly older horizon. Occurrence: McNairy sand member of Rip Although the Ripley species is better pre ley formation, Coopei' pit, near Hollow Rock, served, the essential differences are slight, and Carroll County, Tenn. if a single botanic species may be regarded Genus ALISMAPHYLLUM Berry as varying within the limits of the two it is quite possible that Potamogeton middendorf AHsmaphyllum cretaceum Berry, n. sp. ensis. and P otamogeton ripleyensis may rep Plate III, figure 6 resent the variants of a single botanic species. Leaves ovate-auriculate, approaching or in "\V'"hen fossil leaves from slightly different cipiently sagittate in generJal outline, with en:. horizons show recognizable differences and in tire margins, pointed tip, and slightly decur termGdiate forms are absent from the geologic rent base.. Length about 4.5 centimeters; record, good practice requires that they shall maximum width, in the basal half of the leaf, · be considered distinct species. about 2.8 centimeters. Petiole stout, curved. Between 30 and 40 fossil species have been about 6 millimeters in length, midrib rathe~ referred to this genus, none of which appear broad but flat. The apparent breadth of both to be identical with the present form. They the petiole and the midrib seems to be due to range in age from the ArCtic Emscherian the fl·attening during fossilization of a not through the Eocene, Oligocene, Miocene, and very resistant structure. The secondary ve Pliocene to the Pleistocene-in fact, several nation is obscure in all the four specimens of still existing species are recorded from the this species in the collection; the most that Pleistocene, both in this country and abroad. can be made ou.t are one or two veins diverg The modern species number more than 60 ing from the midrib at an acute angle at or and occur bot11 in the Tropics and more abun near its base and pursuing an upward and ap dantly in the Temperate Zone. The· species parently aerodrome course parallel with the generally have a wide range, and many are leaf margins. cosmopolitan, a single 'species commonly ex In so far as its characters 'are discernible tending ovep a great· many degrees of lati- this species should be referred to the family m I:Ieer, Oswald, Flora fossilis arctica, vol. 7, p. ~9, pl. 55, Alismaceae. Rather than give it· a too d~ figs. 23, 24, 1883. cisive generic position I have referred it to the 2s Ben·y, E. W., U. S. Geol. Sut-vey Prof. Paper 84, p. 27, pl. 4, fig. 6, 1914. genus Alisrnaphyllum, which was proposed by 36 THE FLORA OF THE RIPLEY FORMATION me 29 for leaves from the latest Lower Creta- This characteristic little leaf, unfortunately ceous (Patapsco formation) of Virginia. Ge- known only from the single specimen figured, neric names for fossil forms supposed to be- is an indubitable member of the Dioscoreaceae long to this family are plentiful; thus· A lis- and is seemingly one of the most 1nodern ele 'lnacites Saporta 30 contains a doubtful Lower ments in the Ripley flora, but this is due, I Cretaceous form from Portugal, a second from ! believe, to the scarcity of this family in the the Dakota sandstone, which is totally differ- fossil record throughout the Tertiary rather ent from this Ripley species, and two or three than to its late evolution, as the present species Oligocene and ~1iocene species of Europe. tends clearly to indicate. Alismaphyllites was proposed by l{nowlton 31 Fossil forms referable to the family Dios in 1918 for a form from the lower Eocene coreaceae are almost unknown. Lesquereux, (R'aton formation) of southeastern Colorado. i.t is true, described Dioscorea? cretacea,32 The genus Alisma has had referred to it a fron1 the Dakota sandstone of l{ansas, but doubtful forin from the Upper Cretaceous this forin is entirely valueless as evidence for (Atane beds) of western Greenland and sev- th~ presence of the famil}7 in the Upper~ eral Tertiary species, as well as still existing Cretaceous. The present species, however, forms occurring in the Pleistocene. furnishes conclusive proof of the presence of The family contains about 75 existing spe- this family toward the end of Cretaceous cies segregated into about a dozen genera, all time. It is very similar to existing species of which Plate III, figure 7 Leaves as preserved linear, presumably lanceolate above and sheathing below, 4 ..5 to Leaves small, ovate, ·widest in the basal half .6 centimeters in width and preserved without of the leaf. Apex narrowly rounded. Base ·any diminution in width for a length of 50 broadly rounded. Margins entire, very centimeters. Texture very coriaceous. Mar- slightly undulate. Substance thin but firm. gins ·entire. Veins sin1ple and parallel, im ~1idrib slightly stouter than the latei·al pri mersed, considerably less than 1 1nillimeter maries, of which there are three on each side, apart. Leaves alike on both surfaces. In the rather evenly spaced and. aerodrome. They hollows between the veins occ.ur rows of small are connected by oblique nervilles, ascending sto·mata with the guard cells all oriented in a from within outward and rarely forking. direction parallel with the veins and equally numerous on both sur•faces of the leaf. Leaf ::9 Berry, m. W., Maryland Geol. Survey, Lower Cretaceous, p. 452, 1911. surface under the microscope ap]_1earing finely 80 Saporta, Gaston de, AnnaJes sci. nat., Bot., 4th ger., vol. striated parallel with the veins. 17, p. 228, 1862.. 31 Knowlton, F. H.,. U. S. Geol. Survey Prof. Paper 101, 38 LeS()uereux, Leo·, Cretaceom: flora, p. 56, pl. 28, fig. 10, p. 286, 1918. 1874. ARECALES 37 Occurrence: Eutaw formation (basal part), ess of eliminating other previously described 2 miles south of I-Iavana, 1-Iale County, Ala. species of Geonom~tes which show the distin Coffee sand men1ber of Eutaw formation, 1 guishing characters that appear 'to be specific. mile north of Beacon, Decatur County, Tenn. As preserved in the Ripley clays the remains Ripley formation (Cusseta sand member). indicate leaves of large size. The accompany Buena Vista, Marion County, Ga. ing illustration was redrawn from· a sketch made from a specimen in place by E. S. Perry Order ARECALES in October, 1919 1 combined with the many fragments collected by Dr. 'Vade. This speci Family ARECACEAE men had a length of 60 centime~ers and a width Genus GEONOlVIITES Visiani of 45 centimeters and represented a leaf at least 100 centimeters in length, if not much Geonomites schimperi Lesquereux longer, and over 50 centimeters in width. The Plate II rays vary greatly in width and are generally united, becoming free marginally and toward Geonom'ites schi·ni<1JePi Lesquereux, Tertiary flora, p. the tip of the leaf. J 16, pl. 10, fig. 1, 1878. As has been previously pointed out, the This species was described as f o1 1·ows b y ....,£renus (}eonomites, which receives its namf' Lesquereux: from its resemblance to the existing genus Geon01na of 'Yilldenow, is more pr.operly con- Frond large; rays flat, obtusely carinate, hnlf sidered to represent 'the undifferentiated ances shen.thing or decurrent to a narrow rachis, obscurely ti'y or the generically indistinguishable fossil nervetl; primnry Ye:ns thici.: or inflated; intermediate nerves few, three or four. representatives of the ~ribe Geonomeae. This tribe includes -a.t least ten genera, three of The specimen represents a portion of an which, all monotypic, are natives of western evidently long and large leaf. The rachis is Africa. The remaining genera, containing 98 narrow, not more tha.n 4 millimeters thick at per cent of the known species, are widely dis the lowest part of the specimen, very slowly tributed in tropical and subtropical America. decreasing in si~e, as at. the top of th~ spe~i- ~1ost of these seven genera are small, contain men, 'vhich is 22 centimeters long, It still ing from one to five existing species each. The measures :3 millimeters. It appears half round only large genus of the ti~ibe is Geonor:7Ut: and striate bitt is distinct only at a few places, · which comprises behveen 80 and 100 existing the rays apparently covering it by their de- species that extend f,rom the Antilles and current or half-sheathing base. The rays are southern J\1exico through Central America, scarcely nar.rowed at the base, enlarging a along the eastern base of the Andes to Bolivia, little above it, and connected by their borders, and along the east coast of South America to thence decreasing upward, soon disjointed: Rio de Janeiro. Their center of distribution . and thus palmato-pinnate; they average in at the present time is in the basin of the Ama width 2.5 centimeters a little above the rachis. zan. They are prevailingly small stemless or 'I'he rays are marked in the middle by a deep short-stemmed palms and are undoubtedly of midrib and are thus subcarinate; they show on American origin. The fossil representatives both sides of it ten to fourteen inflated pri- of the Geonomeae comprise seve.ral Eocene :and mary ·veins, with few intermediate veinlets, Oligocene forms of both North America and only two to four discernible, and even these Europe that have been referred to the genus rarely, after abrasion of the epidermis. The 11/ anicaria of Gaertner, three forms referred to substance of the fronds is thin, membranace- Geon01na, and six forms referred to Geono ous, of a dull-red color, a character which may ndtes. be casual. The present Ripley fossil represents th'j Palm leaves are· notoriously without clearly oldest known OCClirrence not only of the genus defined specific features, especially a·s repre- but of the tribe Geonomeae and the only ce~· sented by the broken remains characteristic of tain pre-Eocene occ"i..ll'rence. . The type of most museum specimens, hence the determina- Geonomites sohimperi came from a somewhat tion of thjs speqies rests la.rgely upon the proc- indefinite locality in Yellowstone Park aud 100913-25-4 • 38 THE FLORA OF T'HE RIPLE.Y FORMATION has doubtfully been referred to the "Lara from the Oligocene of Italy, and a third, like· Inie " of ·that region, which is now assigned to wise referred to Geonorr~Ja, comes fr·om the Mesaverde. If this age determination is sub lower Miocene of Switzerland. . stantiated the Ripley occurrence is probably Although the geologic record is incomplete, of about the same age. The species has abo it would seem to indicate a trop~cal origin for been recorded from beds of uncertain age in the tribe, in later Upper Cretaceous time, a Contra Costa County, Calif., but this record spread northward far into western North may well be doubted, for it is not substan America during the ·early Eocene, and an in tiated by either figures or preserved speci- vasion of southern Europe, possibly extending FIGURE 4.-IOOstoratipn of Geon.omites visia.nii Berry, one-eighteenth natural size mens. In addition to the present Cretaceous to tropical Africa, during the middle Eocene species either Geonoma or Geonomites, or or slightly later. An opinion as to whether both, are recorded from the early Eocene of the tribe migrated to the Eastern Hemisphere Colorado (Raton and Denver formations), across the Tropics or in higher latitudes is Wyoming, New Mexico, and Texas. On the not warranted in the present state of our other· hand, the earliest known Europear1 knowledge. form, the type of the genus, comes from thl3 Geonomites has not yet been found in th P lower Lutetian of Italy. A _second species, extensive Eocene floras fro~ the shores· of the referred to Geonoma by its describer, comPS Mississippi embayment which I have de- • JUGLANDALES 39 ·scribed, ·but in view of its presence in that Class DICOTYLEDONAE region toward the end o:f Upper Cretaceous Series CHORIPETALAE time it should eventually be found in the early Eocene of that region. Order JUGLANDALES The accompanying attempt. at a restora Family JUGLANDACEAE tion of the basal Eocene Geonomites visianii · Berry ,33 o:f western Texas (fig. 4), not only Genus JUGLANS Linne ·"'ill answer for GeonoTnites schimperi, which Juglans similis Knowlton differs in general aspect merely by the split ting a part of the rays distad, but it also serves Ju,glans similis Knowlton, U. S. Geol. Survey Prof. to indicate the general habit o:f the numerous Paper 101, p. 255, pl. 36, fig. 2, 1917. existing species of Geonoma. Somewhat ques This species, which is represented by several tionable fragments of Geonomites schimperi specimens in the Ripley · fonnation, is de· occur in the shell marl at Coon Creek. scribed as follows by Knowlton: Occurrence : MeN airy sand member of Ripley formation, Perry place, Henry County ; I.;eafiets membranaceous in texture, ovate-lanceolate in outline, abruptly w~dge-shaped at base, pro 2lf2 miles south o:f Miftlin, ChestBr County, longed above into an acuminate apex ; margin entire; Coon Creek, l\1cN airy County, Tenn. midrib thin, straight; secondaries ten or twelve pairs, thin, alternate, slightly curved upward, camptodrome, Genus SABALITES Saporta and curving just within the border to join the one next above; finer nervation obsolete. Sabalite~ sp. Berry This fine little species is well represente9- by the example figurep. This is narrowly ovate-lanceolate Sabalites sp. Berry, U. S. Geol. Survey Prof. Paper in shape, being about 7.5 centimeters in length and 112, p. 72, pl. 11, fig. 1, 1919. 3.25 centimeters in width in the middle. The nerva- . Fragments of leaves o:f a large :fan palm tion is well shown in the figure. are present in the basal Ripley formation of This species is of the type of certain of the leaflets Be~ ton and MeN airy counties, Tenn. They often referred t This species is exceedingly well marked and synthesis of many fragments, the distal teeth is entirely distinct from previously described on each side are aqui~ine conical lobes, on the forms. It resembles closely some of.the leaves left side two outward-directed teeth are sepa of our existing Oarmptonia pe1·egrina (Linne) rated by wide openly rounded sinuses, and the Coulter. · It is also much like some of the proximal and sligh~ly more produced tooth is European Tertiary forms, about which so produced as an ascending conical lobe; on the much controvetsy raged in tim·es past as to right side the penultimate tooth is but slightly whether they were myricaceous or protea developed and directed outward, the next be ceous.36 For example, some· of the forms of low is a conical ascending lobe, below which Oomptonia vindobonensis (Ettings):lausen) is a normal outward-directed tooth, followed Berry are similar to the present species. A by a proximal and slightly more produced somewhat similar form from the Bohemian tooth. 'l'he petiole is stout and about 5 milli Cretaceous is described by V elenovsky 37 as· meters in length; the midrib is stout, slightly Dryandra cretacea, and another from the Cre flexuous, and prominent on the lower surface taceous of Transylvania by Unger 38 as Oomp of the leaf. The secondaries are thin; usually tonites antiqrulus. These are both generically a craspedodrome secondary proceeds from the distinct from the present species, as shown by midrib at varying angles to the tips of the teeth their characteristic habit. 111 yrica ripleyensis or lobes; the rmnainder, which are numerous, is not especially close to any of the other Rip are camptodrome, or unite at large angles with ley species of Myrica, Jrf. wadii being more an infra1narginal vein or hem, which is· ap elongated and slenderer, with conspicuously parently the modification of a more regular different marginal characters. camptodrome vena.tion due to the formation Occurrence: Ripley for1nation (MeNairy of the marginal sinuses. The ultimate areola- · sand member), Camden-Paris road, 13 miles tion can not be made out. northwest of Camden, Henry County, Tenn. ; The species is not uncommon but fragmen 2Jj2 miles southwest of Selmer, McNairy tary, and small fragments suggest small lo County, Tenn. bate lea.ves of Quercus, although the venation is not that of Quercus but suggestive of Myrica Myrica cooperensis Berry, n. sp. :or· some member of the Proteaceae-for ex amp-le, Lomatia. General considerations, it Plate VI, figure 1 seems to me, rule out the latter possibility, so 0 that I am constrained to adopt 11! yrica, in spite Leaves oblong-lanceolate, of medium size, of the abundance and diversity of Myrica with an acuminate apex and a narrowly cu types that this flora is thus made to exhibit.. neate base, the 1nargins irregularly toothed or There is little likelihood of confusing the pres lobed and the texture coriaceous. Length, ent unique type either with its associates in the about 9 centimeters; 1naximum width, at a Ripley flora or with other American forms. point where the marginal teeth may become On the other hand, in both size and peculiar lobate, about 2 centimeters. The characteristic marginai lobing, it is somewhat like a form features of these species is furnished by the from the Senonian of Germany described by peculiar and v.a.riable character of the margins, Hosius and Von der Marck39 as Qiwrcus hiera~ which range from remote. outward-directed ciifolia. As the name used by Hosius and Von teeth separated by nearly equilateral rounded der Marek is one in. the literature commonly sinuses to variants in which one or more of the attributed to Debey, it is probable that some ·teeth are directed ·upward or finally produced what similar forms occur in the Cretaceous of as -~ascending narrow · conical pointed lobes~ Aix-la -Chapelle. Thus in the specimen figured, which is a The value of careful descriptions and precise ao See BerTy, E. W., Living and fossil species of Comp figuring of fossil leaves is well illustrated by tonia: Am. Naturalist, vol. 40, pp. 485-520, pis. 1-4, 1906. the unsatisfactory objects described by Hosius s1 Velenovsky, Josef, Die Flora der bohmischen Kreide formation, pt. 2, p. 1,. pl. 1, figs. 1-5, 1883. and Von der Marek. I should .like to call the as Unger, Franz, Ueber einige fossile Pflanzenreste nus Sie benbtlrgen und Ungaro : K. Akad. Wiss. Wien Sitzungsber., 39 Palaeontograpbica, Band 26, p. 42,. pl. 31, figs. 85-88, Band '51, pt: 1, p. 2, pl. l, fig. ·1, 1865. . L880 MYRICALES 43 Ripley species 111 yrica hieraciifolia but am pre mation of New Jersey, but this record was vented by the possibility that the German form based on an. erroneous identification. Heer re may have to be ·transferred to 111 yrica. The de ferred the type material to the genus Quercus, sirability of more precise comparison· has a but it lacks any positive diagnostic features of bearing upon correlation. Mqreover, there are that genus and is on the other hand identical a much larger number of botanists interested in essential characters with numerous species in the origin an¢!. distribution of the oaks. of frl y1'ica. The tendency o"f paleobotanists to Formerly almost any lobate leaf was referred see resemblances to Quercus in fossil leaves to Quercus, often upon scanty evidence. One with a toothed or lobate margin irrespective would not find fault with the older workers if of other and ·more important foliar characters. they ·had left figures accurate enough for a is greatly to be depi·ecated. judgment. If, !as seems probable, Quercus and The present species is distinguishable with Castanea are derived fr01n Dryophyllum, the· difficulty from the Patoot forms ·named Quer lobate forms older than DT1.jophyllun1 are open cus marioni by Heer,40 which similarly is to to suspicion, and with scientific descriptions . be referred to 111 yrica. Both the Patoot flora and figures it would be possible to get some and that of the Ripley furnish a galaxy of idea of the lines of evolution. mostly small, toothed-margin fonns .that are Occurrence: l\tlcNairy sand member of Rip very liable to confusion unless their characters ley formation, Cooper pit, near I-Iollow Rock: are carefully analyzed. Carroll County, Tenn. Occurrence : MeN airy sand n1ember of Ripley formation,. Cooper pit, near Hollow Myri~a johnstrupi (He~r) Berry Rock, Carroll County, Tenn. Querc~ts johnstntp·i Heer, Flora fossilis arctica, vol. Myrica minor Berry, n. sp. 7, p. 24, pl. 56, figs. 7-12a, 1883 [not Newberry, 1896]. Plate VI, figures 2, 3 Knowlton, in Stanton and Martin, Geol. Soc. .Leaves small, elliptical, with an acute tip America Bull., vol. 16, p. 408, 1905. and a cuneate or somewhat cordate base. Mar Leaves lanceolate to ovate, with a bluntly gins with a few large 0 crenate-serrate teeth. pointed apex and abroadly cuneate and slight Texture subcoriaceous. Length 1.75 to 2 cen ly decurrent base. Texture subcoriaceous. timeters; maximum width, in the middle pa1t Length from 3 centimeters to 7 millimeters; of the leaf, 8 to 13 millimeters. Petiole stout, maximum width, at or· below the middle, from curved, 3 to 5 1nillimeters in length. Midrib 1.25 to 2.5 centimeters. Margins with sOine stout, prominent on the under surface of the what variously developed coarsely dentate . leaf. Secondaries stout; four or five pairs teeth, which are generally 1l.1ore prominent in diverge from the midrib" at acute angles and the larger leaves. Petiole stout, of undeter ascend in rather straight courses to the tips of mined length. Midrib stout, prominent on the the marginal· ·teeth ( craspedodrome) . Ter- under sur.face of the leaf: Secondaries thin but tiaries obsolete. . prominent, six to eight subopposite to alternate This little species is somewhat variable, the pairs diverging from the midrib at angles of extremes being figured. In the broader forms about 45°, pursuing a rather straight subparal , the base becomes cordate and the basal sec lel ascending course, and terminating at apices ondaries diverge: at wider angles. Super of the marginal teeth, hence craspedodrome. ficially these tiny leaves suggest the juvenile Tertiaries thin, percurrent, usually obsolete by · leaves of the associated 11/yrica o1·nata Berry, immersion. b~t their long petioles negative this sugges- Tllis species is. abundant in the Patoot beds .tion. Although -represented by only three of Greenland and was well illustrated by Heer: specimens, the species is present at the two It has been recorded from the Upper Creta principal outcrops where Ripley plants have ceous of Chignik Bay, Alaska, but this mate been collected in Tennessee. · rial has not yet been described or figured. It Occurrence: MeN airy sand member. of Rip . i:=; sparingly represented by characteristic ley formation, Perry place, Henry County, specimens in the Ripley of Tennessee. It was 0 ' Heer, Of>wald, Flora fossilis arctlca, vol. 7, p. 23, pl. 56, recorded by Newberry from the Raritan for- figs. 1-6. 1883. 44 THE FLORA OF THE RIPLEY FORMATION and Cooper pit, near Hollow .Rock, Carroll Coulter, but the latter are shortly pointed or Comity, Tenn. rounded distad, with the 1nargins lobed rather than toothed. Somewhat similar Terti'ary Myrica ornata Berry, n. sp. forms in both Europe and North An1erica are Plate VI, figures 4, 5 0 omptonia vindob onensis ( Ettingsha usen) Ovate-lanceolate leaves of medium size and Berry and Omnptonia insignis (Lesquereux) striking appearance, varying somew h'at in size Berry. Attention should also be called to the and proportions, generally widest below the leaves from the Patoot beds of Greenland middle, tapering upward to the more or less which Heer 42 described as Quercus johnstrupi extended acuminate tip and dow:invard to the and which demonstrate the presence in Green broadly or narrowly cuneate base. Margins land of a very similar and congeneric type also entire for the basal one-fourth their length, present in the Ripley flora and referred to lily their upper three-fourths divided. into large rica by n1e in the present work. aquiline serrJate cuspidate teeth, which decrease Occurrence: JVIcNairy sand member of Rip regularly in size above the middle. Teeth ley formation, Cooper pit, near Hollow Rock, from seven to ten on each side, varying from Carroll County, Tenn. a crenate.-serrate form to the elegant type re produced photographically, in which the Myrka. torreyi Lesquereux prmninent teeth are slightly awned. Lea.f ·plate VI, figures 6, 7 substance thick and polished, subcoriaceous. Length from 7 to 9 centimeters; maximum Myrica torreyi Lesquereu~, U. S. Geol. and Geog. Sur width from 1.9 to 3.1 centime(ers. Petiole vey Terr. Anp. Rept. for 1872, p. 392, 1873 ; U. S. Geol. Survey Terr. Rept., vol. 7 (Tertiary short and stout, curved, about 2 millimeters in flora) , p." 129, pl. 16, figs. 3-10, 1878. length. Midrib stout, prominent on the under Ward, U. S. Geol. Survey Sixth Ann. Rept., p. 551, side of the leaf; seven to ten subopposite p'airs pl. 40, fig. 4, 1886; U. S. Geol. Survey Bull. 37, diverge from the midrib at acute angles of p. 32, pl. 14, fig. 5, 1887. about 35°, pursue a but slightly curved ascend Knowlton, U. S. Geol. Survey Bull. 163, p. 34, pl. 6, figs. 1-3, 1900; U. S. Geol. Survey Prof. ing course, and terminate in the marginal teeth Paper 98, p. 90, pl. 17, fig. 7 ; p. 336, pl. 86, fig. 1, ( craspedodrome). ·The tertiaries are im- 1916; Prof. Paper 101, p. 256, pl. 37, figs. 2-4, mersed in the leaf substance. They are thin and 1918. form arches along the margins and are percur- Cockerell, Colorado Univ. Studies, vol. 7, p. 150. . rent between the secondaries. The areolation 1910. is fine and indistinctly seen and is not that of The material from Tennessee is narrower Querc'us, a genus which might possibly be sug :1nd more prominently toothed than the type gested by the general form of the leaf. material but appears identical with the west This handsome species is a characteristic ern forms referred to this species by !(nowlton. element in the Ripley flora and is not uncom . The type was collected near the Cretaceous mon at the Oarroll ·County locality. It is Eocene boundary at Black Buttes, vVyo., and readily distinguished from the numerous asso the species has been recorded from a consider ciated species of llfyrica. Among previously able number of localities and horizons in the described. fossil forms it is somewhat sugges Cretaceous' of ""\Vyoming, Colorado, and New tive of. Myrica parrvula Heer, of the Raritan Mexico but has not hitherto been found in the formation of New Jersey and the Patoot beds Coastal Plain Cretaceous. It is common at of Greenland, which l redescribed as Oomp the Carroll County locality in Tennessee· but tonia 1nicrophylla (Heer) Berry.41 That has not been found in the collections from form is sm'aller, however, with fewer and ob: IIenry County. tuse teeth, and lacks the' ~cuminate tip. lily Occun~ence: MeN airy sand member of Rip rica ornata also resembles the juvenile leaves ley formation, Cooper pit, near I-Iollow Rock, of the exist~ng Oomptonia peregrina (Linne) Carroll County, Tenn. 41 Berry, E.. W., Am. Naturalist, vol. 40,. p. 508, pl. 4, figs. 42 Heer, Oswald, Flora fossilis arctica, vol. 7, p. 24, pl. ;,r,, 1, 3, 4, 1906. fig>'. 7-J:!, 1t:.s~:. • MYRICALES 45 Myrica torreyi obtusata Berry, n. var. low and undulate or distantly and obtusely toothed above. Texture coriaceous. Petiole Plate VI, figure 8 and midrib fairly stout and prominent on the Leaves relatively small, linear, with a blunt underside of the leaf. Secondaries thin, im- t.ip and a m1rrowly ctineate decurrent base. . mersed, often obsolete, branching from the ~1arg:ins dentate, with an occasional tooth as midrib at rather h~rge angles, comparatively suming a serrate form. Length about 6.5 straight, abruptly camptodrome, forming a centimeters; maximum width about 5 milli pseudo:marginal vein,' wh!ch is especially meters. Petiole stout, curved, 7 millimeters prominent in the Ripley material and which in length. 1\1idrib stout, prominent, usually gives off a short and inconspicuouos brnnch to curved. Secondaries thin but well marked, the marginal teeth. diverging from the midrib at angles· of over This striking species was described by me 45°, alternately craspedodrome to the 1narginal from specimens found in the Magothy forma teeth and abruptly camptodrome, the latter tion of New Jersey and was subsequently re forming with the branches from them to the corded from the Black Creek formation in craspedodrome veins a pseudomarginal vein. South Carolina. At both of these localities Tertiaries thin, forming an open polygonal the matrix is rather coarse, and the detailed areolation. characteristics of .the species come out much The general form and especially the blunt more clearly in the present specimens, pre tip show a decided resemblance to the fossil s~rved in the fine plastic Ripley clay. The leaves commonly referred to the genus species is well marln~d and abundant in the Banksia, but the venation is not that of Hipley deposits of Henry County but has. not Banlcsia and is of the type commonly referred yet been found elsewhere in Tennessee or in to ~1yrica and exactly similar to that shown in the western Cretaceous. the conte.mporaneous forms.referred to 11/yrica Occurrence : ~1cN airy sand member of Rip to1·reyi. The new variety differs from 111 yrica ley formation, Perry place, Henry County. to?'?'eyi in its s1nall size, narrow form, and Tenn. blunt tip. l\1arginal characters are rather va riable among the forms that have been referred Myrica brittoniana obtusata Berry, n. var. to illyrica tor1··eyi. Those from the Cretaceous Plate V, figures 1, 2 of Tennessee 'that are considered to represent This variety differs from the type in its that species have n1ore prominent and 1nore relatively shorter and broader form and in its closely spaced teeth than this new variety. conspicuously rounded tip. Length from 10 Occurrence: lHcN airy sand 1nember of Rip to 12 centimeters; maximum width from 2 to ley formation, Cooper pit, near Hollow Rock, 2.5 centimeters. The midrib "is prominent Carroll County, Tenn. below and channeled on the upper surface of the leaf. The secondaries are more numerous, Myrica brittoniana Berry more prominent, and straighter, and the mar~ Plate VI, figure !.l ginal vein is more conspicuous. This variety is n~t abundant and is asso My1··ic(t brUtoniana Berry, Torrey Bot. Club Bull., ciated with the normal form of 11/ yrica b?·it vol. 32, p. 46, 1905; New Jersey Geol. Survey Ann. Hept. for 1905, p. 138, 1906; U. S. Geol. toniana and 1nay therefore simply represent Survey Prof. Paper 84, p. 31, pl. 7, figs. 17, abnormal leaves of that species. The more 18, 1914. obtuse tip that has suggested the name for Jlfyriun hec1·ii Berry, Am. Naturalist, vol. 37, p. 68:2, this variety is a feature of slirrht weio-ht but . ~ b ' figs. 7, 8, 1903 [homonym, Boulay, 1887]. the channeled 1nidrib and more prominent Leaves elongate-lanceQlate, 13 to 14 cent~ and slightly different secondary venation seem meters long by 2.7 c~ntimeters in greatest to be characters of diagnostic value. width, which is in the middle part of the leaf. Occurrence: :MeN airy sand l!l~mber of Rip Apex elongated, narrowed, bluntly pointed. ~~y formation, Perry place, Henry County, Base attenuated. Margin e~1tire, or entire be- I enn. 0 46 THE FLORA OF THE RIPLEY FORMATION Order SALICALES about midway between the apex and the base, Family SALICACEAE from 1.75 to 2.5 centin1eters. Petiole missing. Texture subcoriaceous. Margin with regularly Genus SALIX Linne spaced, fairly prominent, nearly straight ser Salix gardneri Knowlton? rate teeth. Midrib stout, prominent on the Plate V, figure 3 lower surface of the leaf. Secondaries thin, ·regularly spaced, about fifteen craspedodrome Salim gardneri Knowlton,. U. s: Geol. Survey Prof. pairs ·branching from the 1nidrib ,a.t angles of Paper 101~ p. 257, pl. 37, fig. 1, 1919. 45° or more, curving regularly upward, sub The present species is described as follows parallel, terminating in the marginal teeth~ by l{nowlton: Tertiaries thin, partly percurrent and partly Leaves linear-lanceolate, broadest about the middle, alternating, joined midway between adjacent thence tapering gradually to the long wedge-shaped secondaries by a zigzag vein. lower portion and in about equal degree to the acumi This well-marked species is represented by nate apex; margin perfectly entire; midrib fairly strong; secondaries numerous, about twelve pairs, several specimens from the MeN airy sand alternate, at an angle of about 45°, very much curvecl member of the Ripley formation of Tennessee. upw·ard, each joining the one next above, the lower None of these are complete, each showing about ones by a series of several loops, the upper ones mak two-thirds of a leaf, but this is enoi1gh to ing themselves · a series of · conspicuous loops ; demonstrate their identity with the European nervilles prominent, approximately at right angles to the midrib. type, which came from the Turonian or Em '.rhis .handsome little species is represented by the scherian of Aix-la-Chapelle, Rhenish Prussia. nearly perfect example figured as well as several Additional European localities are Tannen more or less fragmentary specimens. The. figured berg, Bohemia, and I\:ieslingswalde, Silesia., all specimen is 7 centimeters long and 1 centimeter wide: at about the same horizon and probably in the the others are. slightly smaller. This species is undoubtedly most· closely related to Santonian substage of the Emscherian. Salim plicata Knowlton, of the Vermejo formatioit, Occurrence: Ripley formation (MeN airy with which it agrees closely in size and shape but sand member), big cut on Southern Railway from which it differs in the disposition of the less near Cypress and 2¥2 miles southwest of Sel numerous secondaries; the finer nervation is also a mer, McNairy County; Cooper pit, near Hol more conspicuous feature in the present species. low Rock, Carroll County (rare) , Tenn. The Ripley material is only tentatively identified with this Vermejo species, which is · Dryophyllum protofagus Berry, n. sp. only doubtfully. related to tlw genus Salix. The Ripley material shows a long petiole and Plate V, figures 4, 5 is much like ~ variety of lanceolate members Leaves rel'atively small, alternate in arrange of the Ripley flora but differs from all of these ment, oval, widest in the median region and in venation. ~.bout equally tapering to the acute apex and Occurrence: MeN airy sand member of Rip-· base, the apex slightly more tapering than the ley formation, Perry place, Henry County, base. Margins deiltate. Texture subcoria Tenn. ceous. Petiole short and stout, about 2.5 milli: Order FA GALES meters in length. Midrib stout a.nd prominent Family FAGACEAE on the lower surface of the leaf. Secondaries stout, subopposite to alternate, numerous, sub Genus DRYOPHYLLUM Debey parallel, craspedodrome; they diverge from Dryophyllum gracile Debey the midrib at angles of 45° to 50°, pursue a Dryophyllu.m gracile Debey, Feuilles querciformes nearly straight ascending course, and termi d'Aix-la-Cllapelle, p. 10, figs. 10, 11, 1881. nate in the marginal teeth .. Tertiaries well Berry, Torrey Bot. Club Bull., vol. 43, p. 290, pl. marked, numerous, prevailingly percurrent. 16, fig. 6, 1916; U. S. Geol. Survey Prof. Paper The n1argins are entire· for a variable distance 112, p. 75, pl. 32, fig. 2, 1919. below, and the teeth increase in size somewhat Lanceolate to o.blong-lanceolate, witli a cu toward the tip of the leaf. Length about 3.5 neate base and a gradually narrowed tip. centimeters; maxiinum width about 1.7 centi Length about.12 centimeters; maximum width, meters. .1A;~ 48 .l ,~ •. .c: FLORA OF· THE RIPLBY FORMATION margins, along \vhich they arch. Tertiaries Family MORACEAE comprise arches along the margins \Y hich send Genus ARTOCARPUS Forster short branches into the teeth and percurrent Artocarpus cretacea Berry, n. sp. veins between tlie secondaries which are con nected transversely in th.e median region. · Plates VIII, IX These leaves, which are somewhat suggestive Leaves large, elliptical or broadly ovate, of some modern species of the genus Oordia, deeply divided into long, narrow pinnate seg resemble Oeltis in their general form and mar ments. Estimated length 40 centimeters; gina,! teeth, curved petiole and midrib, slightly 1naximum width about 22 centimeters, in the falcate fo~·m, and secondary and tertiary vena median region. Texture coriaceous. Apex tion. acuminate. Base decurrent. This species is The genus Oeltis comprises shrubs and trees unfortunately based upon fragments, the larg of the warm temperate ai).d tropical portions est of which are shown on Plate IX. These of the Northern I-Iemisphere, with outlying show, however, the· apex, the base, and a me species in the ~1ascarenes, New Caledonia, dian por'tion of the leaf, and I have had in ad Java, and Polynesia. It is divided into sev clition a sketch of a larger median portion eral subgenera, some of which are probably made by E. S. Perry from a specimen in place of generic rank, such as the score o~ tropical in the clay pit. From all this material the res American forms of 11/ mnisia Dumortier. The toration shown on Plate IX, greatly reduced Euceltis section of the genus contai:p.s between in size, has been reconstructed. The reader 30 and 40 existing species of the warmer parta should bear in mind that such features ·as the of North A1nerica and Eurasia. A consider.: actual number of lobes may have been more or able number of fossil species of Oeltis have less or the direction of the lobes below the been described, but the only previously known a pical portion represented by actual specimens form from the Cretaceous with such an im may have been more ascending. Simila.rly the plied relationship is Oeltidophyllu1n p'raeaus sinuses may not all have been so deep as in the trale l{rasser,44 fr01n the Cenomanian of Kun median and basal regions actually represented stadt, ~1oravia. This assignment does not ap by specimens. Aside from these features the pear especially convincing to me, and I would restora.tion is absolutely correct. be inclined to consider the form a .Zizyphus. The species· may now be described in n1ore The oldest certainly determined forms here detail. The leaf is divided into about ten pairs tofore known con1e from the lower and middle · ·of opposite to alternate linear-lanceol ate acumi Oligocene of France. The. genus is well repre nate lobes separated by nearly parallel sinuses sented in the European Miocene, with no less which extend almost to the midrib, the lamina than eight species, and is present at Florissant, in these regions constituting little more than a in the Colorado Rockies, as well as in Wyo narrow wing. The lobes are somewhat irregu ming. The Pliocene species include repre _lar in size and spacing, especiaJly in the apical sentatives in l;>oth Europe and eastern Asia. and basal portions of the leaf; they expand ·The existing European Oeltis au.st1·alis Linne. slightly outward, and some of those in the is found in the Pleistocene of Hungary, and median portion of the leaf, which form an an both 0. occidentalis Linne and 0. mississippi gle of ·about 65 ° with the midrib, · develop a ensis Bose, as well as an extinct species, occur subordinate rounded. lobule on their distal in the Pleistocene of the United States. margins about two-thirds of the distance from Occurrence : ~1cN airy sand member of Rip the base to the tip. The basal lobes are more ley formation, Cooper pit, near I-Iollow -Rock, ascending, and the apical lobes are especially Carroll County, Tenn. unsymmetrical and falcately ascending. _The petiole is very stout, of u:p.known length but 44 Krasse1·, Fridolin. Beitriige zm· Kenntniss der fossilen Kreicleflorn. von Kunstadt in Mllh1·en, p. 18, pl. 6, figs. 8-14, presumably short, with winged margins 1896. formed by the decurrent lateral margins of URTICALES 47 This well-marked small leaf conforms in its Apex acute. Base broadly cuneate. I\1argins characters to the genus Dryophyllum and also entire at base, above which remotely and shows many characters prenuncial of the· coarsely · crenate. Texture subcoriaceous .. genus Fagus of this family, a fact which has Length about 8 centimeters; maximum width suggested the specific name. The character about 3.5 centimeters. Petiole stout, enlarging of the margin and the details of venation are proximad, about 9 millimeters in length. I\1id well shown in the figured enlargement of a rib stout, prominent, curved. Secondaries thin, portion of the leaf. Among previously de subparallel; about eleven. equally spaced pairs scribed fossil forms the present species shows diverge from the midrib at approximately considerable resemblance to a species .from the regular intervals at angles of about 50° to 55°, Patoot beds of Greenland which Heer de- pursue a but slightly curved ascending course, . scribed 43 as Quercus 'lnarioni but which is and terminate in the marginal teeth. Terti probably a 11/ yrica and not a Quercus. aries obsolete. The genus Dryophyllu'ln embraces a large Five Upper Cretaceous species have been re number of well-characterized forms found ferred to the genus Fagus-one from Saxony, throughout the late Upper Cretaceous and one from Nortlnvest Territory (Peace River), early Eocene of the Northern Hemisphere. It and three frmn the Dakota sandstone. None has long been considered, and rightly so, the of these are at all similar to the present species, ancestral stock of Oaslianea, Oastanopsis, nor has the genus been recorded frmn the Up Pasania, and Quercus. Its relationship to the per Cretaceous of Greenland. The exjsting modern genus Fagus has not been so clear, but beeches fohn a compact group of four species the present Ripley species shows plainly that of eastern Asia, southeastern North America, Fagus was undoubtedly derived from the same and Europe. Their segregation indicates a stock. A great many Cretaceous species of fonner rnore extensive distribution, abun Q'!.te'l·cus have been described, largely on insufti- dantly verified by the numerous species of the - cient evidence, and a large . number of the geologic record. earlier fossil oaks of the willow-leafed and Leaves similar to F a,qus 1·ipleyensis have chestnut-oak types are probably not true oaks. been referred to both· Qnercus and Planera but The present distribution of the family is a are in my judgment representative of a differ subject of exceeding interest. The direct an ent stock. cestry of the modern forms is still obscure, and Occurrence: MeN airy ·sand member of Rip much more information is needed regarding ley formation, Cooper pit, near Hollow Rock, . the Tertiary forms, especially of the ·Asiatic Carroll County, Tenn . · region. The general relationship among the Order URTICALES different genera. may be expressed· diagram Family ULMACEAE matically as follows : Genus CELTIS Linne ~ Dryophyllum 1~:~- I ~ ---,___ Celtis cretacea Berry, n. sp. Plate VII, figure 2 Pasania I A ------Leaves of medium size, ovate, with a gradu 1 · //Castanopsis _>~gus ally narrowed acuminate tip and a broadly 1 . N othofagus J · 1 0 astanea ~ rounded base. Margins entire in the rounded Occurrence: MeN airy sand member of Rip basal region, above whi~h they are regularly ley formation, Cooper pit, near Hollow Rock, coarsely serrate. Leaf substance thin. Length Carroll County, Tenn. about . 9 centimeters; maximum width, in the basal half of the ~eaf,. about 3 centimeters. Genus FAGUS Linne Petiole stout, curved, ·expanded at base, about Fagus ripleyensis Berry, n. sp. 1.2 centimeters in length. Midrib stout, Plate V, figure 6; Plate VII, figure 1 curved, prominent on the lower surface of the leaf. Secondaries thin, five ·or six pairs, the Leaves of medium size, widest in the middle basal pair opposite and slightly suprabasilar, and narrowing about equally to both ends. the remainder alternate; they diverge from 43 Heer, Oswald, Flora fossilis arctica, vol. 7, p. 23, pl. 56, the midrib at rather wide angles and curve figs. 1-6, 1ssa. upward, ascending parallel with the lateral ~r URTICALES 49 ~ . 'i!.HT the basal lobe. The midrib is very stout and \Vest to about the same time that it appeared prominent on the under side of the leaf. The in the Cretaeeous record of west Greenland. secondaries or midveins of the lobes are rela . In the existing flora the two score or more tively stout, prominent on the under side; they known species of A rtocarpus are indigenous to conform to the angle and direction of the the southeastern Asiatic region, although some lobes, terminating at the tips of the lobes. of them are cultivated in all tropical•countries. The tertiary venation is exceedingly fine and The breadfruit is found throughout Oceanica largely immersed in the leaf substance. It and was present in Hawaii and the ~1arquesas comprises irregplarly spaced tertiaries, of when those islands were first visited by Euro which some preserve a camptodrome course peans. It was introduced into the vVest Indies but more are lost in the prevailingly quad in 1793. Of the tribe Euartocarpeae, of which rangular and rather open areolation. A.rtocarpus is the largest existing genus, five · The present species is strikingly distinct genera are confined to Centra'! and South from. previously described species of A ?·to America, one is confined to tropical west ca'rpus in its extremely pinnate lobation, de Africa, two to the southeastern Asiatic region, . parting in this respect from both the recent one to Borneo, and one ranges from Japan to and the described fossil forms. If it is cor Australia. Though the geologic history of rectly identified it \Vould se·em tha.t so special A. rtocctrp~ts is only imperfectly known, at least ized a form implies the existence of other and 27 different fossil specie.s have been described. less modified forms in the same or· antecedent The oldest that is certainly identified is a well Cretaceous deposits. The only other fossil marked form based on characteristic leaves and forms that it seems at all necessary to discu.ss parts of the fruit that show the typicaJ surface in the present connection a.re Artocarpus dis ftatures. It was fully described by N athorst 48 secta J(nowlton,45 from the Vermejo formation and came from the Atane beds (Cenomanian) of southeastern Colorado, and Li1iodendron of west Greenland. IIere would belong the Da S'IW'Wii Lesquereux,40 from the Dakota sand"'· kota sandstone form called Li1·iodendron snowii stone of l(ansas. Both·· are clearly distinhiodenrl1·on, the most bizarre fossil forms the Laramie formation, and the genus is widely of which it does not in the least resemble, distributed in the basal Eocene of North and that it should be referred to the. Euarto~ America. , It. continu~d in the Gulf region until enrpeae; at least both Liriodendron snottoii the end of the Oligocene, the latest recorded and Artocarpus dissecta are congeneric, and occurrence being in the sands of the Alum if A1·toca1·p~t8 cretacea is correctly identified Bluff formation at Alum Bluff on Apalachicola then the other two should be referred to the River, western Florida. On the P'acific coast same or an aljjed genus. Similarly Li1·io~ it, is found in deposits in California and Ore dendron quercifoliuni Newberry 47 is open to gon \vhich are referred to the Miocene. In .the suspic.ion and the same interpretation. This :European area a great variety of forms de w·ould trnce bnd~ the ancestral breadfruit in scribed by Principi have been found in lower · North America. both in the East and in the Oligocene ( Sannoisian) of Italy (Liguria). The genus is also represented in the Tongrian ~ 5 Kn()wlton, :n·. H., U. S. Geol. ~urvey Prof. Paper 101, p. of France, the Tortonian of Baden, the Pon ' 267, pl. 42, fig. 6, 1917. 40 • Lcsqncreux, Leo, U. S. Geol. Survey Mon. 17, p. 209, pl. tian of France and Italy,. and the Pliocene of 29, fig-s, 1., 2. 1892. "Newberry, J. S., U. S. Geol. Survey Mon. 26, p. 81, pl. 48 Nathorst, A. G., K. Svenska Vet>Akad. Hand I., Band 24, ·.11. figs. 1-6, 1896. No. l, 10 pp., 1 pl.. 1890. 50 THE FLORA OF THE RIPLEY FORMATION. Italy. It is present in both the Pliocene and stouter midrib. Ficus atavina Heer 51 has a Pleistoreene of the island of Java. somewhat similar venation but is a relatively A rtocarpus is said to be represented by petri much narrower, more elongated form; the fied wood in the Oligocene of the island of An associated Ficus crassipes Heer and Ficus tigua, and it was evidently a member of the krausiana Heer have a decidedly different A1nerican flora from the Upper Cretaceous venation. until late in the Tertiary, although, like the Occurrence : MeNairy sand member of the genera Oinnaanomrum, Nipa, and Phoenix, it is Ripley formation, Cooper pit, near Hollow not represented in post-Pleistocene American Rock, ·Carroll County, Tenn .. floras. · Ficus krausiana Beer Occurrence: McNairy sand member of Rip ley formation, Perry place, Henry County, Ficus krausiana Heer, Schweizer. paleont. Gesell. and Cooper pit, near Hollow Rock, Carroll Neue Denkschr., vol. 23. p. 15.. Pl. 5, figs. 3-6, 1869. County, Tenn. Fric, Natui·wiss. Landesdurchforschung Bohmen Archiv, vol. 4, pp. 18, 94, 1878. Genus FICUS Linne Lesquereux, U. S. Geol. Survey Mon. 17, p. 81, Ficus ripleyana Berry, n. sp. pl. 1, fig. 5, 1892. . Hollick, U. S. Geol. Survey Mon. 50, p. 58, pl. 9, Plate VII, figure 6 fig. 9; pl. 10, figs. 1-3, 1907. Leaves lanceolate, widest in the middle Berry, U. S. Geol. Survey Prof. Paper 84, pp. 38, . 110, pl. 11, figs. 4-7 ; pl. 19, fig. 4, 1914; region and tapering almost equally upward Maryland Geol. Survey, Upper Cretaceous, p. and downward, the 'apex· being slightly more 823, pl. 59, ·fig. 1, 1916. extended and acuminate. Base acute, decur Ficus beclcwith'ii Lesquereux, Cretaceous and Tertiary rent. Margins entire, somewhat undulate. floras, p. 46, pl. 16, fig. 5 ; pl. 17, figs. 3, 4, 1883. Texture coriaceous. · Length about 15 centi Ficus snspecta Velenovsky, Die Flora der bohmischen Kreideformation, pt. 4, p. 10, pl. 5, figs. 6, 9, meters; maximum width about 3.75 centi ·1885. meters. Petiole very stout and round, about 1 centimeter in length. Midrib very stout, Leaves generally large, ovate-lanceolate, . pr01ninent on the under surface of the leaf. broadest at or below the middle. Apex and Secondaries thin but well m'arked, nur..1erous, base acutely pointed, the apex often extended diverging from the midrib at angles of ;tbout and attenuated. Petiole and midrib stout. 65°-70° at regular intervals of about 4 I.nilli- Seconda.ries regular, open, thin, ascending, ·meters and abruptly camptodrome in the mar:- camptodrome, branching from ·the midrib at ginal region. Areolation typical of the mod-: I angles of 45° or more. Length about 17 centi- ern fig leaves of the Ficus benjamine.a and meters. · Ficus elastica type. This well-known Up.per Cretaceous species The general form of this ~pecies is very was described originally from specimens similar to. that of the mostly widespread and found in the ·cenomanian of Moravia, and it abundant Upper Cretaceous species such as has been subsequently recorded .from both the Ficu$ crassipes Heer, F. krausiana Heer, and Cenomanian a.nd Turonian of Bohemia. It F. atavina Heer, with which specimens failing occurs at a large number of American locali to show venation 'are Iia.ble to be confused. ties. In the vYest it occurs in the Dakota Ficus georgiana Berry ,4o from the Ripley of sandstone; in the East it is common from Georgia, is relatively much wider, with a more Marthas Vineyard to Alabama and is present open venation ·and more truncatedly rounded between these limits in Ma.ryland, North Caro-. base. Ficus mata1..oanensis Berry,5o of the Mata- lina, South Carolina, and Georgia. These oc wan formation in the New Jersey area, has a currences are all in beds of Magothy age or similar venation but 'a more linear form and younger. In both North and South Carolina . Ficus fruits ·are associated with this species~ ~ Berry, E. W., U. S. · Geol. Survey Prof. Paper 84, p. 111, pl. 20, fig. 1, 1914. 111 Heer, Oswald, Flora fossilis arctica, vol. 6, pt. 2, p. 69, 150 Berry, E'. W., Torrey Bot. Club Bull.;. vol. 38, p. 399, pl. pl. 11, figs. ub, 7b, 8b; pl. 17, fig. 8b; pl. 19, fig. 1b; pl. 20, 19, fig. 3, 19'11. figs. 1, 2, 1882. URTICALES 51 but whether they are to be r~ferred to it or form differs decidedly frmn the present one in . to some of the other rather numerous species its ascending camptodrome secondaries. of Ficus that occur at the same localities can . Occurrence: MeN airy sand member of Rip not be determined. · ley formation, Pen·y place, Henry Cou~ty, It is especially abundant in the Middendorf Tenn. arkose member of the Black Creek formation Ficus crassi,pes (Beer) Beer in South Carolina and occurs in the lower Eutaw h1 both Georgia and Alabama, and in Plate VII, figure 5 the Coffee sand member of the Eutaw fm;ma Proteoides cmssipes Heer, Flora fossilis arctica, vol. tion in western Tennessee. It. has not hereto 3, pt. 2, p. 110, pl. 31, figs. 6-8, 1874. fore been recorded from a horizon as young as Ficus crassipes Heer, idem, vol. 6, pt. 2, p. 70, pl. 17, the Ripley, nor is it at all common in the Me fig. 9a ; pl. 24, figs. 1, 2, 1882. Lesquereux, U. S. Geol. Survey Mon. 17, p. 79, N airy sand member. In the; West it does not pl. 13, fig. 3, 1892. 0 appear to have been present in post-Dakota Berry, U. S. Geol. Survey Prof. Pap~r 84, pp. 37, time. 110, pl. 10, fig. 4; pl. 12, figs. 8-10, 1914; Mary Occurrence: MeN airy sand member of Rip- land Geol. Survey, Upper Cretaceous, p. 821, pl. ley formation, Cooper pit, near Hollow Rock, 58, fig. 5 ; pl. 59, figs. 2, 3, 1916. Carroll County, Tenn. Leaves entire, narrowly lanceolate, about equally tapering to the acuminate apex and Ficus obtusa-sessilia Berry, n. sp. base. Length 12 to 20 centimeters; greatest Plate VII, figures 3, 4 width, in middle :part of leaf, 1.8 to 2.5 centi nleters: Texture coriaceous. Midrib stout Spatulate leaves of medium size, widest often extraordinarily so. Secondaries thin ' above the middle, with a broadly rounded tip open, ascending,. camptodrome. ' and a cuneate, eventually truncate and clasp This species was described originally from ing base, with a thin but somewhat polished material obtained in the Atane beds of west surface and evenly rounded entire margins. ern Greenland, and the first rather fra .. 52 THE FLORA OF THE RIPLEY FOR.MATION Ficus celtifolius Berry undulate; petiole short, thick; nervation strong, the midrib straight or at most slightly fiexuose, with from Plate X, figure 1 three to six pairs of strong secondaries, the lowest l.i'icus celtifoli1ts Berry, U. S. Geol. Survey Prof. Paper pair arising at the base of the blade, of the same size 84, p. "37' pl. 12, fig. 4, 1914. as the midrib and together producing a three-ribbed effect; the lowest pair of secondaries (or ribs) with Leaves small, entire, elliptical-ovate, with an from. three· to as many as ten tertiary branches on the obtusely pointed apex and a broadly roundt:d outside, the lowest of these, 'in the larger leaves, · base. J\1argins entire. Texture subcoriaceous. nearly at right angles to the midrib, with a numb.er of quaternary branches on the lower side, which sup· Petioie relativ.ely stout, of variable length. ply the extreme basal portion of the blade ; the sec J\1idrib stout. Secondaries stout, remote, ondaries on .the midrib are usually remote, alternate, \ three or four pairs, the lo.wer pair opposite, or subopposite, and. sometimes· with a few branches diverging from the m:idrib at its base; or on the outside; all nervation camptodrome and arch just above and functioning as lateral pri ing just inside the margin; nervilles numerous, ·strong, mainly unbroken ; finer nervation producing an abun maries, the rest alternate and less straight. dance of irregularly quadrangular areolae. All brancho from the midrib at angles of about In all these leaves the nervation is essentially the 45° and are camptodrome some distance from same. 'l'hus, regardless of size, all are three-ribbed the n1argin. Tertiaries distinct, the inner from the very base of the blade, and all show the same camptodrome secondaries or their branches, with transverse~ the peripheral camptodrome in uni the same strong, mainly unbroken nerville~. As the form broad arches. leaves become more deeply heart-shaped at base the This handsome species was first found in th~ secondary and tertiary nerves which supply the basal 1\fidclenclorf arkose member of the Black Creek portion of the blade become more and more branched. formation in South Carolina, where it is rare. Occurrence: MeN airy sand 1nember of Rip The South Carolina form was 5.5 centimeters ley formation, Cooper pit, near Hollow Rock. in length and 3.7 centimeters in maximum Carroll County, Tenn. (not common). width in the basal half of the leaf and had a petiole 6 millimeters in length. The Ripley · Ficus cooperensis Berry, n. sp. form, which is not common in the collections, Plate X, figure 2 is obviously identical specifically with the type Leaves fairly large, ovate, with full rounded but is somewhat smaller, being just over 4 margins, widest below the 1niddle, with a centimeters in length and 3.1 centimeters in pointed, not prolonged, apex and a ·full bas~ maximum width, with a petiole 9 millimeters rounded in general form but ·slightly decur · in length. It is very much like the smallest rent at the petiole. Texture coriaceous. leaves in the Vermejo flora which ICnowlton Length about 11 centimeters; maximum width refers to Ficus leei. 53 about 5.5 centimeters. Petiole very stout, Occurrence : MeN airy sand member of Rip curved, about 1.5 centimeters in length. J\1id ley formation, Cooper pit, near Hollow Rock, rib very stout, prominent on the lower SUl'·, Carroll County. Tenn. face of the. leaf, slightly curveq. Secondaries Ficus leei Knowlton thin but prominent, about nine opposite to alternate subparallel pairs eli verge from the F'ic1ts leei Knowlton, U. S. Geol. Survey Prof. Paper 1 98, p. 338, pl. 90, fig. 2, 1916 ; Prof. Paper 101, , midrib at ·ang~es of about 45° and are campto p. 261, pl. 39, figs. 1-6; pl. 40, figs. 1, 2, 1917 .. drome in the n1arginal region. Tertiaries thin but well marked, percurrent. This species, which is exceedingly common This belongs to a type of ovate fig leaf in the Vermejo formation of southeastern which is con.;non in both recent and fossil Colorado and northeastern New J\1exico, is de floras, one in wh:ich the basal secondaries tend scrib~d by J(now lton as follows : to be slightly more prominent than those above Leaves very thick and fieshy in texture, the nerva them and eventually lead to forms ]jke Ficns tion strongly impressed; shape broadly elliptical, ovate schimpe1-i Lesquereux or Ficus psmtdopopulus elliptical, or sometimes nearly orbicular the base from rounded and truncate to deeply cordate-auriculate, the Lesquereux, found in the lower Eocene . of apex abruptly rounded and obtuse or sometimes slightly this same region. pointed; margin entire or occasionally very slightly Occurrence: :MeN airy sand meml:;er of Ripley formation, Cooper pit, near Hollow Rock~ r;a See, for example,. U. S. Geol. Survey Prof. Paper 101, pl. 39, fig. 3, 1917. Ca;rroll County, Tenn. PLATANALES, RANALES 53 Ficus georgiana Berry prominent midrib, more ascending secondaries, and less prominent areolation. It is a common Ji'ictts ge01'(Jiana, Berry, U. S. Geol. Survey Prof. Paper 84, p. 111, pl. 20, fig. 1, 1914. form. in the MeN airy sand member, being much more abundantly represented in the col This species is represented in the collections lections than any other of the Ficus species from the Ripley of western Tennessee only by present. It may be matched by a considerable the lower half of a single leaf and is therefore number of existing species of Ficus and also included with some hesitation, although the suggests comparisons with the lower Eocm1e rounded base and decurrent margins appear Ficus wilcoxensis Berry, which may represent characteristic. The spercies was described by a vVilcox descendant of the present form. It me in 1914 as follows: is named from its occurrence in Carroll Leaf ovate-lanceolate in outline, gradually tapering County. to an acute point. Proximately the leaf is broadly Occurrence: J\1cN airy sand member of Rip and abruptly rounded to a point within 2 to 3 milli ley formation, Cooper pit, near Hollow Rock, meters of the midrib, thence decurring as a narrow wing 1 millimeter or less in width and preserved for Carroll County, Tenn. a ·distance of 1 centimeter. Length about 17 centi meters. Greatest width, which is near the base of the Order PLATANALES lea:r., 4.5 centimeters. Midrib fairly stout. Margins Family PLATANACEAE entire. Secondaries thin, parallel, and .numerous, branching from the midrib at angles of about 50° and Genus PLATANUS Linne curving upward distall~, camptodrome. Occurrence: J\1cN airy sand member of Rip Platanus ripleyensis Berry ley formation, Perry place, Henry County, Platanus 1·ipleuensis Berry, U. S. Geol. Survey Prof. Tenn. Cusseta sand member of Ripley forma Paper 112, p. 84, pl. 32, fig. 6, 1919. tion, near Buena Vista, Marion County, Ga. This species was described in the publica (the type locality for this species). tion cited above. So far as known it is con Fi,cus carrollensis Berry, n. sp. fined to the marine Ripley near Eufaula, Bar bour County, Ala. Plate X, figure 3 Platanus sp. Berry Leaves of variable size, lanceolate, with an Plata?ws sp. Berry, U. S. Geol. Survey Prof. Paper acuminate tip and a gradually narrowed acu 112, p. 85, pl. 31, figs. 7, 8, 1919. minate base. Margins entire. Texture cori aceous. Length from 9 to 16 centimeters; This form, of which only incomplete mate maximum width from 2 to 3 centimeters. rial is ... known, was described and figured in Petiole short and stout. Midrib stout, promi the publication cited above. It occurs in the nent on the lo,ver surface of the leaf. Sec .Ripley of eastern Alabama. In. several par ondaries thin, numerous, subparallel, diverg~ ticulars it suggests the genus Oredneria of ing from the midrib at angles of about 50° at Zenke~ · intervals of about 2· millimeters, pursuing a Occurrence.: Ripley forri1ation, Cowikee rather straight ascending course to the mar Creek, Barbour County, Ala. ginal region, where their tips are joined by Order RANALES arches subparallel with the margins. Areola tion thin, reticulate, mostly obsolete. Family MAGNOLIACEAE These leaves have a form and venation such Genus LIRIODENDRON Linne as is frequently. r0ferred to Eucalyptus .or llfy'rtoJJhyllu·m but is more strictly compar Liriodendron laramiense Ward able to tha.t of lanceolate species of figs of the Plate XI, figure 13; Plate XXIII, figure 6 Ficus elastica type. The present species shows considerable resemblance to Ficus laurophylla Li1·iodena1·on lamntiense Ward, U. S. Geol. Survey Lesquereux, of the Dakota sandstone of the Bull. 37, p. 102, pl. 48, fig. 2, 1887. West and the J\1agothy formation of the At Leaves relatively large, rectangular, with lantic Coastal Plain and may be descended nearly vertical sides, a wide truncate base and from that species. It differs in having a less a deeply emarginate· apex, giving it a, butter- 54 THE FLORA OF THE RIPLEY FORMATION fly-like form. Apical lobes inequilaterally L. alatun~ figured by l(nmvlton 55 and differs rounded. Length 9 to 10 centimeters; maxi from all of these in lacking the conspictJous mum width 9 to 10 centimeters; petiole stout, alate petiole, which is so striking a feature of not alate, its total length unknown. Midrib that species. stout, straight, and 'prominent on the lower sur The past history of Liriodendron is exceed face of the leaf. Secondaries pinnate, five or six ingly interesting and has been commented upon 6 pairs, stout and prominent proximad, diverg at length by me in a number of earlier papers. 5 ing from the midrib at angles of about 60°; At the present time 28 fossil species have been camptodrome, brachiodrome; the suprabasilar recorded, some of which are of doubtful value, pair, which are subopposite, curve· upward although the n1ajority are convincing evidence more than the others and give off numerom; of the past variation, abundance, and wide camptodrmne branches from their ·outer sides; distribution of the genus. Twenty-three of eventually the secondaries unite with those ad these fossil species are recorded from the Up jacent to form the characteristic Liriodendron per Cretaceous ; all of these except one are loops some distance within the margin, which from North America, and this one, Lirioden is bordered all around by smaller loops of the dron schwarzii Richter,57 is an exceedingly third and fourth order. N ervilles traversing doubtful identification from the Senonian of the larger areas percurrent and straight or Saxony, which I have no hesitation in disre curved, or forking and inosculating in the garding ·entirely. The only other recorded middle region; those within the festoons re Cretaceous occurrence of Liriodendron in solving into an areolation of regular quadrate Europe is the identification by Engelhardt 58 of or rectangular meshes. Texture subcoriaceous. Liriodendron meeldi I-Ieer in the Cenomanian . This is a very characteristic type of Lir-io of Nie.derschoena, Saxony, and this same species dendron leaf in size, outline, and venation. has been reported by Kurtz 59 from the Upper The present specimens from Tennessee are Cretaceous of Argentina. As neither of these almost exact counterparts of· the less perfect occurrences has been figured it is impossible to type, which came from the Montana group at speak with absolute assurance, but the fact Point of Rocks, Wyo. The type failed to show that Liriodendron meekii is one of the least the deeply emarginate sinus reaching two well-characterized species of Liriodendron and fifths of the distance to the base, but as the one to which a great variety of leguminous two agree exactly in all the details observable leaflets have been referred by former students there can be no doubt that they are identical leads me to disregard also these supposed . or that the Point of Rocks form had a similar .Cretaceous o'ccurrences in Europe and South emarginate apex. " America as having been based upon incorrect The form of Liriodendron laramiense is one identifications. · often approximated by lea,;es of the modern This conclusion is to some extent rendered Liriodendron tulipifera in localized situations, more certain by the great variety of perfectly such as are supposed to have a phylogenetic characteristic forms of Liriodendron in North bearing. That · it . does not represent such a America at that time. Here they are found . variant of a normally lobate Liriodendron of from the Atlantic to the Pacific in great Upper Cretaceous age is substantiated by the abundance. In the present Atlf~:ntic Coastal exact agreement in form between these two Plain region they extend fron1 Long Island to occurrences over 1,000 miles apart and the en Alabama, Tennesse~, Arkansas, and Texas. In tire absence of any other forms of Liriodendron us Knowlton, 11'. H., U. S. Geol. Sur:vey Bull. 163, p. 13, pl. either at these localities or in beds of the same 1, fig. 6, 1900; U. S. Geol. Survey Prof. Paper 101, p. 269, age in the vicinity. In general outline Lirio pl. 48, fig. 3, 1918. so Berry, E. W., Torrey Bot. Club Bull., vol. 28, pp. dendron lararniense is somewhat suggestive 493-498, pis. 41, 42, 1901; Torreya, vol. 1, pp. 105-107, pis. of Liriodendron alatum Hollick,54 of the Belly 1, 2, 1901 ; vol. 2·, pp. 33-37, pis. 1, 2, 1902 ; voL 3, pp. 129-132, 4 figs., 1903 ; Bot. Gaz., vol. 3~, pp. 44-63; 3 figs., River and V ermejo formations, particularly 1902. . . the large oblong type of that species. It is less 57 Richter, P. B., Beitrage ~ur Flora der oberen Kreide Quedlinburgs, pt. 1, p. 6, pl. 1, fig. 10, 1905. like the smaller and more rounded specimens of 58 Engelhardt, Hermann, Naturwiss. Gesell. Isis in Dres den Abh. 7, p. 100, 1891. M Torrey Bot. Club Bull., vol. 21, p. 467, pl. 220, 1894. 00 Kurtz, Federico, Mus. I,a 'Plata Rev., vol. 10, p. 53, 1903. RANALES 55 the vV est they are found in ICansas, Nebraska, back to a rather complete . understand~ng of Colorado, Montana, Utah, Iowa, and on Van this history being the lack of information couver Island. Most of these are clearly de about the Cretaceous floras of Asia, a knowl fined and readily recognized distinct species, edge of which would probably greatly modi although it is possible that some of the eleven· fy our p1~esent conceptions. All the undoubted species recorded from a small area in the Da Cretaceous records are seen to be North kota sandstone of l(ansas and Nebraska may ·American, and all the Tertiary records are represent merely. variations of fewer botanic Arctic and Eurasian. The results of species. In North American Tertiary strata Pleistocene glaciation restricted the range of Liriodendron has not yet been recognized, al the American form and its Chinese .variety to though it would seem . that this genus must the r·egions shown. Although many paleo have been present, for the existing species geographers have questioned a la.nQ bridge is very common in the Pleistocene of southeast across the north Atlantic and see in the suc ern North America. In Europe, on the other cessive convergences and divergences of the hand, Liriodend'ron appears to have been abun terrestrial faunas and floras of Europe and dant, especially in the later Tertiary, as its max- North America a succession of interchanges FrounE -5.-Sketch map of the world summarizing the geologic history of Liriodendron. Cretaceous occurrences, lined area within heavy ellipse; Tertiary occurrences,. lin.ed areas within latitudinal lines; existing distribution in solid black. Arrow indi cates the ··Upper Cretaceous-Eocene path of migration · · imum occurrences are in the Pliocene and come and separations in the Bering Strait region, rjght down to the first glacial deposits, in the presenc~ of an early Tertiary Liriodendron which fruits and seeds were found which the in west Greenland, Icehtnd, the British Isles, R.eids 60 were unable to distinguish from the and central Europe would seem to indicate existing American species. Presumably the . a north Atlantic path of migration as shown genus was as abundant in Asia, as undoubted on the accompanying map. leaves are recorded by Schmalhausen 61 from Occurrence : MeN airy sand member o·f. the Altai region of central Asia in beds of R.ipley formation, Perry place, Henry County~ probably Pliocene age. ';['his form undoubt Tenn. edly represents the stock from which ·came the Genus MAGNOLIA Linne existing Chinese variety. Magnolia capellinii Beer The geologic history of the genus is graphi Magnolia capellinii Heer, Phyllites cretac~es du cally summarized on the accompanying sketch Nebraska, p. 21, pl. 3, figs. 5, 6; 1866; Flora map of the world (fig. 5), the only draw- fossilis arctica, vol. 3, Abt. 2, p. 115, pl. 33, figs. 1-4, 1874 ; idem, vol. 6, Abt. 2, p. 90, pl. 24, oo Reid, C. and: E. :M:., Pliocene flora of the Dutch-Prussian figs. 3-5, pl. 25, figs. 1-3, pl. 45, fig. 1, 1882. border, p. 03, pl. 8, figs. 1-5, 1915. Velenovsky, Die. Flora der bohmischen Kreide 61 Schmnlhausen, J ., Pnlaeontographicu, vol. 33, p. 211, pl. 21, figs. 20, 21, 1887. formation, pt. 2, p. 20, pl. 7, figs. 8, 9, 1883. 56 THE FLORA OF THE RIPLEY FORMATION Lesquereux, The flora of the Dakota group, p. the fossil species that have been referred to 203, pl. 46, fig. 1,. 1892. illenispe1"17'1ites, its most constant features are Dawson, Roy. Soc. Canada Trans., 1st ser., vol. 11, sec. 4, p. 63, pl. 11, fig. 49 ; pl. 13, fig. 49a, its trilobate form and peltate base, well shown 1894. in the Ripley specimens. A more complete Hollick, New York Acad. Sci. Trans., vol. 12, p. knowledge of the species n1ust await better ma 234, pl. 6, fig. 6, 1893; New York Bot. Garden terial, b:ut it is clearly distinct from other mem Bull., VOl. 3, p. 413, pl. 78, fig. 3, 1904; The bers of Upper Cretaceous floras and is rather Cretaceous flora of southern New York and New England, p. 63, pl. 17, figs. 3, 4, 1907. readily recognized even in f~·agmentary speci Berry, Torrey Bot. Club Bull.,· vol. 31, p. 76, pl. mens. The type came from McBrides Ford, in 3, fig. 3, 1904 ;~New Jersey Geol. Survey Ann. Chattahoochee County, Ga., and the present Rept. for 1905, p. 138, 1906 ; Torrey Bot. Club occurrence demonstrates for the species a geo Bull., vol. 34, p. 195, pl. 12, figs. 4, 5, 1907; U. logic range through the Eutaw and a consider S. Geol. Survey Prof. Paper 112, p. 89, pl. 18, fig. 1; pl. 32, fig. 7, 1919. able portion of the Ripley formation. Nothing Magnolia sp. Berry, Johns Hopkins Univ. Circ., new clo·sely similar fr01n the Upper Cretaceous ser., No. 7, p. 81, 1907. floras of other regions has been described. This species has been repeatedly described Occurrence: MeNairy sand member of Rip and figured in recent years. It has been found ley formation, Perry place, Henry County, at but a single locality in the Ripley and is not Tenn. altogether typical, so that there is a possibility Order PAP A VERALES that the Ripley material represents a different species. Family CAPPARIDACEAE Occurrence: Eutaw formation (basal beds), Genus CAPPARIS Linne McBrides Ford, Chattahoochee County, Ga. Capparis proeocenica Berry, n. sp. Ripley formation (J\1cNairy sand member), 21h miles southwest of Selmer, McNairy County, Plate X, figures 4, 5 Tenn. Graceful leaves of small size, oblanceolate, Family MENISPERMACEAE widest medianly, with a rounded tip and acute decurrent base and entire, evenly rounded mar Genus MENISPERMITES Lesquereux gins. Length from 3.5 to 5 centimeters; maxi Menispermites variabilis Berry mum width from 6 to 11 millimeters. Petiole Menispe1·mites variabilis Berry, U. S. Geol. Survey short and stout, about 2.5 millimeters in length Prof. Paper 84, p. 113, pl. 21, figs. 1-4, 1914. in the sm·aller specimens to 6 millimeters in the This species was discovered in the Eutaw larger. Midrib stout and straight. Second formation of western Georgia and described· in aries thin, numerous, and subparallel; they di the following terms: verge from the n1idrib at angles of about 45° and are camptodrome in the marginal region; Leaves of medium size, trilobate but variable in shape. Length about .9 or 10 centimeters. Greatest the lowest pair are opposite and parallel with width, which is from tip to tip of the laterally di the margins. rected lobes, as much as 16 centimeters in one speci This species presents some features that sug men. Lobes broad and separated by wide, shallow gest the Leguminosae, but on the o'ther hand it sinuses. Margin entire but in some specimens scal loped, the scallops ·rounded and nearly equilateral, is very similar to certain· existing American separated by acute sinuses. Base peltate, broadly species of 0 apparis, as well as to the smaller rounded as in A.spidiophyllum or '':ith a cordate sinus, leaves of Oapparis eocenica Berry,62 which is which does not, however, reach the top of the petiole. a common form in the lower Eocene C\Vilcox Primaries stout, three or four in number, palmately divergent from the peltate base, three of them gen group) of the Mississippi embayment region. erally equal in caliber. Secondaries slender and nu This. resemblance is so conclusive that the pres merous, some apparently camptodrome and others ent form is regarded as ancestral to the Wilcox craspedodrome. form, a feature emphasized in the name. The This peculiar species is represented by frag genus has not heretofore been recognized in mentary remains which contribute nothing to beds ·as old as the .Upper Cretaceous, and the ward a better understanding of its characters. 62 Berry, E. W., U. S. Geol. Survey Prof. Paper 91, p. 218. Exceedingly variable .in outline, as are most of pl. 44, figs. 1-3: pl. 52. fig. 5, 1916. ROSALES 57 present species is not common in the collections with acute apex and base. and entire slightly on which this paper is based. undulate margins. Texture firm and scarious. The genus OapzJaris embraces more than Venation comprising about ten subparallel 100 existing species of shrubs or small trees, aerodrome veins, the central ones a pproxi chiefly tropical, and although found also in mated to simulate a midrib, the lateral ones the Eastern I-Iemisphere most of the species equidistant. They are connected by many thin occur in the American Tropics, particuhtrly oblique forking nm·villes. in Central and South America. OazJparis This unique form, represented by two speci domingensis is a small Antillean tree, and its mens, finds its closest modern analogue in cer leaves, much like this Cretaceous species, are· tain species of Acacia in which the leaves are rather sma1ler than those of most of the other reduced to phyllodes, as, f<;>r ex~ul.1ple, in , members of the genus. Several of the 'iV est Acacia longifolia 'iVilldenow. The genus Indian forms-for example, Oapparis fer Acacia comprises between 450 and 500 existing 'l''tt,r}'inea Linne, 0. mnygdalina Lamarck, and species of tropical and subtropical regions, 0. oynozJhalloz;ho'ra Linne-are shrubs or small l~trgely oriental in their distribution. The sec- . trees of the strand flora, 0. fen"ttginea being tion Phyllodineae, to which the present fossil especially common in such an environment. appears referable, includes more than 300 ex The fossil species is somewhat similar to a isting species confined to Australia and Ocean .. form from the Pliocene of Bolivia described ica. Similar fossil species are extremely rare, 03 by, Engelhardt as Oappa1is 1n1tltine1· althoi..1gh several are known from the Tertia~·y ·vis, which is compared with the existing Uap of both Europe and Americ~. JJa?··is ang'Ltstifolia Humboldt, Bonpland, and Occurrence : MeN airy sand member of 1\:unth, of southern ~iexico; Oapparis jacobinae Ripley formation, Perry place, I-Ienry County, ~ioricaud, of .Brazil; and OapzJaris longifolia, Tenn. fr~m the Antilles, all forms comparable with Genus MIMOSITES Bower bank OapJJm•is p1·oeoce11:ica. Unger 04 many ·years ago described OapzJmis ogygia, from the mid:. Mimosites cooperensi~:~ Berry, n. sp. clle Miocene of Parschlug, Styria, but Schirn Plate X, figure 7 per 05 referred that species to the Phaseoleae. In addition Schenk has described petrified ma Petiolulate lanceolate leaflets of somewhat terial from the Tertiary of Egypt as 0 apJJa'ri variable size, with pointed tip and narrower doxylon, and F. von MUller has described two cuneate base, slightly inequilateral. 1\iargins or three species of fn1its of the genera Dieune entire. Texture subcoriaceous. Petiolule nncl PlesiocapzJa1·is fr01i1 the late Tertiary o-f stout, about 2 millimeters in length. :Midrib Australia. Plesiocapparis is said to be most stout, curved, prominent. Secondaries thin closely related to the section. Busbeclcia of but well marked, numerous, camptodrome. OapzJaris. · Length from 1.4 to 2.5 centimeters; maxinllnn Occurrence: 1\icNairy sand member of Rip width, slightly above the middle, from 3.5 to ley formation, Perry place, I-Ienry County, 5 millimeters. Tenn. This well-marked species is fom~d at both Order ROSALES the principal plant-bearing Ripley outcrops Family MIMOSACEAE in western Tennessee. It appears to possess Genus ACACIAPHYLLITES Berry the characters,of the genus i11imosites, which Acaciaphyllites cretaceum Berry, n. sp. is so extensively developed in the lower Eocene Plate X, figure 6 ('Vilcox group) of the ~iississippi embay Inequilateral sessile falcate-lanceolate phyl ment region, and may be regarded as the late Upper Cretaceous prototype of those forms. loclelike leaflets, widest below the middle1 Occurrence : ~1cN airy sand member of oa Engelhardt, Hermann, Naturwiss. Gesell. Isis in Dresdet•. Ripley formation, Perry place, Henry County: Sitzungsb.er., 1894, p. 7, pl. 1, fig. 18. o., Ungm·, li'ranz, Genera. et species planta.rum fossilium, p Cooper pit, near IIollow Rock, Carroll County: 44R. 1850. . 2lj miles south of ~1iffiin, Chester County, oG Schimpel', W. P., Pul(iontologie v~getale, vol. 3, p. ::>30, 2 1874. Tenn. 58 THE FLORA OF. THE RIPLEY FORMATION Family ,CAESALPINIACEAE Genus BAUHINIA Linne · " Genus CAESALPINITES Saporta Bauhinia ripleyensis Berry Caesalpinites perryensis· Berry, n. sp. Bauhinia ripleyensis Berry, Torrey Bot. Club Bull., Plate X, figure 8 vol. 43, p. 294,· pl. 16, fig. 1, 1916; U. S. Geol. Survey Prof. Paper 112, p. 100, pl. 23, fig. 7, Narrowly obovate, markedly inequilateral 1919. sessile leaflets with entir~ margins, · rounded apex, cuneate base, and decidedlv coriaceous Leaves of medium size, more or less bilobate texture. Length about 3 centim~ters; maxi but much less deeply divided than in the pre mum width, in the median. region, about 9 ceding species, obovate. Length along the millimeters. · lVIidrib very stout, prominent, n1idrib 4.5 centimeters; from apex of lobes to curved. Secondaries few, remote, irregular, base 6.7 centimeters; width across upper part stout, and camptodrome. Tertiary areolation of the leaf 5.5 centimeters. Apical sinus open, well marked, forming a fine isodiametric: mesh. extending ·about one-fourth of the distance to This is a well-marked species, clearly dif ward the base of the leaf, its margins at the tip ferentiated. from associated forms and those of the midrib :forming an angle of about 90°, previously described by the peculiar venation. curving slightly upward and then conspicu It is -represented in the present collections by ously .outward to the pointed tips of the lobes, a single specimen and counterpart. which are directed laterally. Outer margins Occurrence : J\1cN airy sand membe~· of of the leaf full and rounded, becoming straight Ripley :formation, Perry place, Henry County, toward the broadly cuneate base. Midrib of Tenn. medium size. Lateral primaries ~ranching from the base at angles with the n1idrib of Caesalpinites ripleyensis Berry, n. sp. about 25°, of medium size, curving upward and Plate XI, figure 6 then outward, and running to the tips of the . Small, elliptical, inequilateral, sessile leaf lobes.· They give off :four or five camptodrome lets, with broadly rounded, distinctly apicu secondaries on the ol).tside and two or· three on late tip and cuneate base. M~rgins entire. the inside. The midrib, in its upper half, also Texture subcoriaceous. Length. about 1.8 gives off one or two secondaries on each side. centimeters; maximum width, . in median Leaf substance somewhat coriaceous. · region, about 7.5 milli;meters. Midrib stout, This species, which is sparingly represented prominent, straight. Secondaries stout, five or in the argillaceous greensand inarls along six widely spaced paii·s, dive.rging :from the Cowikee Creek and is associated with shallow midrib at wide angles, and camptodrome by a water or estuarine mollusks of the Ripley for single sweeping arch. mation, is markedly distinct from any other Like the other Ripley leaflets of ·the . described species of B auhinia. It is much Leguminosae alliance, this species is rare in smaller and less deeply divided than Bauhinia the present collections. It resembles somewhat gigamtea Newberry or B auhinia alabamen.sis C ae8alpi~ia middendorfensis Berry, 66 of the Berry and is much less ornate in character. It. Mjddendorf arkose n1ember of tl~e Black is, on the. other hand, much larger than B au Creek :formation in South Carolina, but is en hinia marylandica Berry, of the Magothy :for tirely distinct from that species and may be mation in the Maryhnd area. It differs from regarded as the prototype of . the numerous all these American Cretaceous species in its -forms of this type in the lower Eocene (Wil Dointed, outward-directed lobes but is not un cox group) flora of the ~{ississippi embay.ment like a· number of existing species of this genus. region. Occurrence: Cusseta sand member of Ripley Occurrence: MeNairy sand member of formation, right bank of Cowikee Creek, one Ripley :formation, Perry place, I-Ienry County, eighth of a mile above its mouth, Barbour Tenn. ·County, Ala.; McNairy sand member of Rip ley formation, 2;-2 miles southwest of Selmer, 00 Berry, m. W., U. S. Geol. Survey Prof. Paper 84, p. 46, pl. 10, fig. 7, 1914. MeN airy County, Tenn. ROSALES 59 Family P APILION ACEAE as a mucronate point. Textt1re coriaceous. Length about 3 to 3·.75 centimeters; maximum Genus DALBERGIA Linne (son) width about 8 or 9 1nilliineters; slightly in Dalbergia cretacea Berry, n. sp. equilateral. Leaflets practically sessile, the short and stout petiolule being not over 1 milli Plate XI, figure 3 ·meter in length. Midrib straight, very stout Leaflets apparently sessile, inequilateral, and prominent. Secondaiies numerous, thin, ovate-emargina.te, widest above the middle. immersed in the leaf substance and visible only Apex widely and openly emarginate, with ob under favorable illumination; they diverge tusely rounded ears and a nearly straight from the midrib at angles of about 45° and sided angular sinus, forming an angle o'f about are camptodrome and scarcely more prominent 105°. Base cuneate, markedly in-equilateral. than the fine tertiary areolation. Margins entire, full and rounded. Texture This relatively long and narrow species is coriaceous. ~Edrib stout, prominent, mark readily distinguished from the associated spe edly curved proximad. Sec01~daries thin but cies of Dalbergia and from previously de well m'arked; four or five pairs diverge from scribed species by its retuse apiculate tip. It midrib at angles of about 45°, somewhat irreg is represented in the collections by two speci ularly spaced and camptodrome. Tertiaries mens and may be ancestral to the lower thin but well marked, forming a very fine iso Eocene species Dalbergia eocenica Berry,69 of diametric areolation of quadrangular or polyg-. this ~arne region. onal meshes. Length about 2.75 centimeters; Occurrence: ~icNairy sand n1ember of Rip maximum width about 1.6 centimeters. ley formation, Perry place, Henry County, This well-m'arked form is not abundant in Tenn. · the collections. It is largei· and more cori aceous than the associated Dalbe1·gia p?·ewil Dalbergia prewilcoxiana Berry, n. sp. coxiana Berry and much broader, more in equilateral, and with a more prominent vena Plate XI, figure 5 tion than Dalbergia pe1"1'yana Berry. It is Leaflets petiolulate, narrowly ovate-emargi much like some of the Greenland Cretaceous nate, only slightly inequilateral, with an emar forms that fleer erroneously referred to LiTio ginate apex and somewhat dissimilar rounded 67 dendron ttneelcii, 'vhich are the same as the ears. Base ·narrow cuneate, slightly inequilat types of fleer's 0 olutea p1·imordialis of the eral. Margins entire, full and evenly rounded. 68 Atane beds, but it is unlike the forms from Texture subcoriaceous. Length about 13.5 the Ullited States that are commonly referred millimeters; maximum width, above the to Oolutea p1imordialis. These all occur at middle, about 5.5 millimeters. Petiolule stout, somewhat e'arlier horizons, and in the absence relatively long (about 3 millimeters). Midrib of comparative material from Greenland the very stout and prominent, slightly curved. problem of revision must be left open, although S_econdaries thin, immersed; five or six campto it seems doubtful if any of these forms repre drome pairs diverge from midrib at acute sent the genus Ooluteg_. angles. Areolation fine, immersed. Occurrence: McNairy sand member of Rip This tiny species is almost identicai with and ley formation, Perry place, Henry County, undoubtedly ancestral to the Eocene embay Tenn. · ment species Dalbergia w-i.lcoxiana,70 a fact in Dalbergia perryana Berry, n. sp. dicated by the name chosen for it. It differs from Dalbergia wilcoxiana in its more promi Plate XI, figure 4 nent ears and in having the venation n1ore Narrow leaflets, spatulate, widest near the completely immersed.. It is readily distin tip, the rather straight margins gradually ap gt1ished fron1 the other Dalbergias in the proaching to form the ·narrowly cuneate base. Ripley formation by its general form and Apex retuse~ the midrib extending in the sinus smaller size. Not common in collections. 07 fleer, Oswald, Flora fossllls arctica, vol. 6, pt. 2, pl. 22, ail Berry, E. W., U. S. Geol. Survey Prof. Paper 91, p. 245, figs. 4, 6, 1882. · pl. 53, figs. 1, 2,. 1916. 68 Idem, p. 99, pl. 27, figs. 7-11; pl. 43, figs. 7, 8. 10 Idem, p. 246·, pl. 53, fig. 7 ; pl. '54, figs., 1, 2. 60 THE FLORA OF THE RIPLEY FOR.MATION Occurrence: MeN airy sand member of Rip late on the narrower side. Texture relativeiy ley formation, Perry place, Henry County, coriaceous. l\1idrib relatively stout. Secon Tenn. daries not differentiated fron1 the transversely LEGUMINOSAE (position uncertain) elongated areolation. Length about 1.65 centi meters; n1axinium width, in upper part of leaf Genus LEGUMINOSITES Bowerbank let, about 5 millimeters. Leguminosites canavalioides Berry This is a rare form of no decided generic af Leguntinosites camavalioide8 Berry, Maryland Geol. finities but differing from the other leaflets Survey, Upper Cretaceous, p. 842, pl. 76, fig. 6, of this alliance represented in the Ripley flora. 1916. Occurrence: l\1cN airy sand member of Rip Leaves compound, probably trifoliate. Leaf ley formation, Perry place, Henry County, lets large, elliptical, \Y.ith a rounded apex and Tenn. base. Length about 7 centimeters; maximum width, in the middle part of the leaflet, about Leguminosites carrollensis Berry, n. sp. G centimeters. l\1argins entire. 'Texture sub Plate XI, figure 1 coriaceous. Petiolule wanting. Midrib stout, becomi1ig attenuated distad. Secondaries nu Small oval or elliptical leaflets with entire merous, thin, camptoclrome, about ten pairs, eli margins, rounded tip, and slightly reduced verging from the midrib at angles of about base. Sessile and with a midrib which is pro 55°. Tertiary areolation ·papiliona~eous, nounced proxi1nad and immersed distad, as mostly immersed. is the balance of the venation. Length about This species represents a ·leguminous leaflet 10 millimeters; maximmn width about 7 to 8 of unknown generi•c affinity l}amed from its re millimeters. semblance to the leaflets of the existing species This form is of unknown relationship among Canavalia, which number about a dozen, of the . the numerous sjmilar leaflets in the Legumino · Tropics of both hemispheres. In the lo\ver sa e. It is readily· distinguishable from the Eocene of southeastern North America there is other members of this alliance found in the an undoubted species of Oanavalia very close Ripley. to the existing Oanavalia obtusifolia (Lamarck) Occurrence: MeN airy sand member of Rip De Canclolle, a common vV est Indian strand ley formation, Cooper pit, near Hollow Rock, plant. Carroll County, Tenn. . This species was described recently from specimens found in the Magothy formation of Genus GLEDITSIOPHYLL UM Berry Maryland. \Vhat appears to be the same Gleditsiophyllum preovatum Berry, n. sp. species was apparently mo~·e common along the Hipley coast. The Ripley leaves are more Plate XI, figures 11, 12 truly elliptical and not quite so ·broad. The Small, markedly inequilateral, obovate, ses texture is subcoriaceous, and the typically le sile leaflets, with full, entire 1nargins, rounded, guminous areolakion is \Veil shown in some of slightly apiculate tip, aJJ!l decidedly inequilat the specimens. The species may well have been eral .decurrent base. Texture coriaceous for ancestral to the lower Eocene forms from the so small a form. Midrib stout,' expanded Mississippi embayment which have been re proximad, attenuated distad. Secondaries nu fen·ed directly to 0 anava:lia. merous, thin, ascending, inosculating, campto Occurrence: MeN airy sand member of Rip drome. Length about 1.75 centimeter; maxi ley· formation, Cooper pit, near Hollow Rock, mum width, above the middle, about 7 milli Carroll County, and Perry place, Henry meters. CoHnty, Tenn. This characteristic small form is similar in Leguminosites perryensis Berry, n. sp. venation to Gleditsiophyllwn proeocen2cun!J Berry, with which it is associated. It is siini Plate XI, figure 2 hiT to and may be regarded as the prototype of Leaflets small, sessile, nearly equilateral, Gleditsiophyllttm . ovat?.tm Berry,n which, spatulate, with an obtusely pointed tip and t\ 71 Berry, E. W., U. S. ·Oeol. Sm·vey Prof. Pa·pet· 91, p. 23!t cuneate base. Margins entire, slightly undu- pl. 51, fig. 1, 1916. ' GERANIALES 61 though of the same size and proportions, has aristate process. Base acute, slightly inequi the secondaries more individual. G. ovatum is lateral. Margins entire, slightly unsymn1et a 1nember of the lower Eocene or Wilcox flora rical. Texture subcoriaceous. Length about in the Mississippi embayment region. 4.5 centimeters; maximum width about 6.5 Occ.urrence : MeN airy sand member of Rip millimeters. Petiolule short and stout, decid ley formation, Perry place, Henry County, edly curved, about 3.5 millimeters in length. Tenn. Midrib stout, curved. Secondaries numerous, thin, immersed, camptodrome. Tertiaries ob Gleditsiophyllum proeocenicum Berry, n. sp. solete. Plate XI, figures 8-10 A wel.l-m'arked form conspicuous among the Ripley species of the Leguminosae alliance by Small, ovate, inequilateral, sessile leaflets, reason of its 1nodern appearance, lanceolate with a bluntly pointed apex and a somewhat shape, and a~·istate tip. Not common in the wider obtuse base. Margins entire, full, round collections. Closely comparable with a va ed. Somewhat chartaceous in texture. Length riety of modern species of Jltfi'lnosa, Oassia, about 2.75 centimeters; maximum width, in the Oaesalpinia, and Gleditsia. medial region, about 7.5 millimeters. Midrib Occurrence: McNairy sand ·member of Rip thin, curved. Second-aries thin but distinct, ley formation, Cooper. pit, near Hollow Rock, numerous, ascending, inosculating and form Carroll County, Tenn. ing inequilateral camptodrome meshes. This well-marked small species resembles a Order GERANIALES host of existing and fossil leaflets of various Family MELIACEAE genera of the Leguminosae alliance. It is therefore referred to the form genus Gledits.io· Genus· CEDRELA Linne pl~yll'UJm !"ather than to Leguminosites, as it Cedrela prewilcoxiana Berry, n. sp. foreshadows the great display of these types in the lower Eocene of the M.:ississippi embayment Plate XII, figure 1 region. It is readily differentiated from the Leaflets practically sessile, lanceolate and other members of the Ripley flora and may be falcate, widest in the middle and about equ'ally considered the prototype in that floraof Oledit pointed at both ends, with regularly entire siophyllum eocenicWJn Berry,72 which is so margins. Texture subcoriaceous. Length abundant in the lower Eocene Wilcox flora in about 3.5 centimeters; maximum width about this region. It is not common in the present 5 or 6 millimeters. Base slightly inequilateral. collections. In smne respects it is similar to Midrib relatively stout, expanded proximad. the form from the ~1iddendorf arkose member Secondaries thin, numerous, camptodrome, im of the Black Creek for~1ation in South Caro mersed. lina described as Acaciaphyllites g'revilleoides This species, which is rare in the present Berry.73 collections, is so similar to 0 edrela wilcoxiana Occurrence: ~1cN airy sand member of Rip~ Berry,74 from the lower Eocene (Wilcox . ley formation, Perry place, Henry County, group) of the Mississippi embayment region, Tenn. as to suggest the ancestral relationship im plied in the name chosen for the Ripley species. Gleditsiophyllum aristatum Berry, n. sp. Its distinguishing characters are its somewhat Plate XI, figure 7 more linear form and expanded midrib at the ·sessile base, the Wilcox species being petiolu Leaflets lanceolate, falcate, petiolulate, acu late. minate, widest below the middle and nan·owed Occurrence : MeN airy sand member of Rip upward, the midrib produced as a conspicuous ley form'ation, Cooper· pit, near Hollow Rock, 7a Bel'l'y, E. W., U. S. Geol. Survey Prof. Paper 91, p. 238, Carroll County, Tenn. pl. 46, 1916. 7s Bet'l'y, E. W., U. S. Geol. Survey Prof. Paper 84, p. 45, 7' Berry, E. W.,, U. S. Geol. Survey Prof. Paper 91, p. 253. J}l. D, figs. 9·, 10, 1914. pl. 56,. fig. 1, 1916. 100913-25---5 62 THE FLORA OF ·THE RIPLEY FORMATION Family EUPHORBIACEAE verge from the n1idrib at angles of about 70° Genus EUPHORBIOPHYLLUM Ettingshausen and are camptodrome in the marginal region. Tertiaries obsolete. Euphorbiophyllum petiolatum Berry, n. sp. This is a relatively shorter and wider form Plate XII, figures 2, 3 with more prominent venation than Euphor- · Anar·omeda novaecaesareae Berry, U. S. Geol. Sur bioph.ylltum antiquu1n Saporta and Marion~ vey Prof. Paper 84, p. 130 (in part), 1914 (not differing from that species as well as from E. Hollick) ; Prof. Paper 112, p. 130, pl. 30, figs. cre.taceum Berry in its obovate form and undu 1, 2, 1919. late margins. Like the other two species of Small obovate or incised leaves with.rounded Euphorbiophyllurn of the Ripley flora, it is apex and decurrent base .. Margins entire ex-. not a common form in the present collections. cept where leaves are incised by deep acute It is very much like E. vetus Saporta,76 from sinus reaching to the midrib. Texture coria the Lutetian of France. ceous. Length about 2 centimeters; maximum Occurrence: MeN airy sand member of Rip width about 6 to 8 millimeters. Petiole stout, ley formation, Perry place, Henry County, long and curved, about 1 centimeter in length. Tenn. Midrib very stGmt, prominent. Secondaries thin, camptodrome. Euphorbiophyllum cre~ceum ·Berry, n. sp. This form has been previously found in both Plate XII, figure 6 the Eutaw and the Ripley formations and Spatulate. leaves of medium size, with a heretofore identified with Andromeda novae broadly rounded tip and a narrowly cuneate 5 caesar·eae Hollick/ although attention was base. Margins entire. Texture coriaceous. called to its obtuse form and sometimes incised Length about 6.5 centimeters; maximum margin. It resembles some of the variants of width, in the upper part of the leaf~ about 1.8 that species but not the normallanceolate form centimeters. Midrib very stout and promi with numerous ascending secondaries. The nent. Secondaries thin, camptodrome, diverg normal type has not been found in deposits ing at wide angles. Tertiaries obsolete. later than the Tuscaloosa formation in the This species is similar to Euphorbiophyllum Mississippi embayment region, the later forms antiquum Saporta and ~iarion but much, all being obovate or oblanceolate, so that they wider, with a finer areolation and more promi evidently represent a distinct species, which is nent midrib, as well as a more coriaceous tex also much 1nore coriaceous. Similar forms tu~·e. It differs fron1 the associated E. tennes have been referred to the genus Orotono seensis' .Berry in general form and in lacking phyll~tm. the undulate margins and prominent second Occurrence: Eutaw formation, Coffee Bluff, aries of that species. It is not common in the Tenn. Cusseta sand member of Ripley forma present collections. tion, Buena Vista, Ga. MeN airy sand member Occurrence: MeN airy sand member of Rip of Ripley formation, Perry · plftce, Henry ley formation, Perry place, Henry County, County; Cooper pit, near Hollow Rock, Car Tenn. · roll County ; 2V2 miles south of Mifflin, Ches ter County, Tenn. Euphorbiophyllum antiquum Saporta and Marion Plate XII, figure 5 Euphorbiophyllum tennesseensis Berry, n. sp. Euphorbiophyllum antiquum Saporta and· Marion, Plate XII, figure 4 Evolution du regne vegetal, Phanerogames, p. 117, fig. 125 c, 1885. Obovate leaves of medium size, widest in the middle with a rounded apex and p.ointed base. Leaves. spatulate or linear-spatulat~, with a Margins entire, conspicuously un~ulate. Tex rounded or slightly emarginate apex and a. ture coriaceous. Length about 5.5 centimeters; narrowly cuneate decurrent base. Margins en maximum width about 1.5 centimeters. Mid tire. Texture subcoriaceous. Length 5.5 to rib stout, .prominent, and curved. Secondaries 6.5 centimeters; maximum width, above the numerous, thin, prominent; about ten pairs di- middle, .1.2 to 1.4 centimeters. Petiole want- 76 Hollick, Arthur, in Newberry, J .. S., U. S. Geol. Survey 76 Saporta, Gaston de, 'Le monde des plantes avant l'appa· Mon. 2·6, p. 121, pL 42, figs. 9~-12; 28-31, 1896. rition de l'homme,. pl. 48, figs. 1, 2. 1879. SAPINDALES 63 ing. Midrib stout, less prominent than in the Order SAPINDALES associated species, curved. Secondaries nu Family CELASTRACEAE merous, thin, ·diverging at wide angles, camp todrome. Tertiaries obsolete. Genus CELASTROPHYLLUM Goeppert The material from Tennessee can not be dis Celastrophyllum carolinensis Berry tinguished :from that from the Turonian of Bagnols, in southern France, figured by OeZastrophyUum carolinensis Berry, U. S. Geol. Sur Saporta and Marion, although it may well be vey Prof. Paper 84, p. 51, pl. 13, figs. 1-fi, 1914; Prof. Paper 112, p. 109, pl. 24, figs. o, doubted if both represent the same botanic 7, 1919. species and not closely related type$ not dis tinguishable by their foliar characters. These This species was described in 1914 as fol- remains are not uncommon in the Ripley flora, lows: · ' where they are associated with three other Leaves lanceolate in outline, with a pointed. apex species of E1"phorbiophyllum. The genus was and a cuneate base, about 14 centimeters in length by 2.9 centimeters. in· greatest width, which is about established by Ettingshausen, 77 as a :form midway between the apex and the base, tapering genus for leaves of the family Euphorbiaceae. equally in· both directions. Midrib stout, rather flexu A considerable number of Tertiary species ous. Secondaries numerous, thin, branching from have been referred to 'it. These are mostly the midrib at acute angles of 45° or less, curving up frpm European localities but include a form ward, usually camptodrome, a few craspedodrome in from Greenland and one from Brazil. The the upper part of the leaf, sending tertiary branches into the marginal teeth. Margin entire for a short genus has not heretofore been recognized in distance at the base, above which it is crenate or this country, except for a single species re- biconvex, the teeth large and interspersed with . corded by n1e from the lower Eocene of Ten smaller subordinate teeth of the same character. nessee, a. descendant from one of these Ripley Leaf substance thin. forms. A second European Cretaceous species, It is not uncommon in the Ripley, where E. primordiale, from the Cenomanian of some of the smaller specimens are only 7 centi Portugal, was described by Saporta.78 meters in length by 1 centimeter in width. Occurrence: J\1cN airy sand member of They also· show considerable variation in mar- : Ripley formation, Cooper pit, near Hollow ginal characters ; the teeth are sometimes less Rock, Carroll County, Tenn. biconvex and more drawn out, lacking sub· ordinate te('th. It is possible that the Ripley Genus MANIHOTITES Berry forms represent a new variety, but the char- . Manihotites georgiana Berry acters are such that a positive conclusion on this point is not possible; they are certainly Manihotites georgiana Berry, ·Torrey Bot. Club Bull.. affiliated with this species, and the :fact of · vol. 37, p. 507, figs. 1, 2, 1910 ; U. S. Geol. their occurrence in younger beds than those Survey Prof. Paper. 84, p. 114, pls. 22, 23, 24, figs. 4, 5; text figs. 2, 3, 1914; Torrey Bot. from which the type was collected suggests Club Bull., vol. 43,. p. 295, 1916. that they may represent a variety. One feature of unusual interest in connection This remarkable form was recently de- with this species is its great similarity in :form, scribed by me in the papers cited above, ~nd texture, and variability to T ernstroemites rip- the discussion need not be repeated here. leyemis Berry. Superficially the two are in· Occurrence: Eutaw formation, basal beds, distinguishable, but the venation shows con 1\1cBrides Ford, Chattahoochee County, Ga.· stant differences throughout a large amount of Coffee sand member of Eutaw formation, Cof-· material. The numerous thiri ascending sec fee Bluff, Hardin County, Tenn. Cusseta sand· ondaries of Oelastrophyllum aarolinensis are member of Ripley formation, near .Buena replaced in Ternstroemites ripleyensis by :fewer, Vista, Marion County, Ga. ; MeN airy sand stouter camptodrome secondaries at nearly member of Ripley formation 21j miles south 2 right angles to the midrib, so that the two are west of Selmer, McNairy County, Tenn. undoubtedly distinct. It seems probable that 77 Ettlngshausen, Constantin von, K.-k. geol. Reichsanstalt this is another of the SO-called Species of 7 1 53 Abh .. vol. 2, pt. 3, No. 2, p. 7, 8 · I Oelastrophyllum that should be referred. t 0 78 Saporta, Ga.ston de, Flore fossile du Portugal, p. 218, . • . pi. 39, fig. 23, 1894. Ternstroem~tes, but as I have been unable to 64 THE FLORA OF T'HE RIPLEY FORMATION satisfy myself on this point I prefer not to Asia should plant beds of this age be subse make any change of name at present. The type quently discovered. material came from the Middendorf arkose. There are 8 species of Oela:strophyllum re member of the Black Creek formation in South corded from the Patapsco formation, 12 from Carolina, and the species was subsequently re the Tuscaloosa formation, 9 from the Raritan corded frmn the Tuscaloosa formation of Ala formation, 7 from the Dakota sandstone, 5 bama, where it is not com1non. from the Magothy formation, 2 from the Black Occurrence: MeN airy sand member of Rip Creek formation, and 3 from the Middendorf ley formation, Cooper pit, near Hollow Rock, arkose member of the Black Creek formation Carroll County, Tenn. in South Carolina. In Greenland there is one species in the Atane beds and three in the Pa Celastrophyllum ripleyanum Berry, n. sp. toot beds. Plate XII, figures 8-~0 As has been mentioned in a previous pub lication, it seems very probable that a number Leaves of small size, obovat~, with. a blunt of these species of Oelastrophyllu1n should be apex and a cuneate and slightly decurrent base. referred to the family Ternstroemiaceae. Texture subcoriaceous. Length about but pre Others, like the Ripley species, belong in the v:ailingly unde-r 2 centimeters; maximum width, family Celastraceae. at. or above the middle, about but prevailingly under 1 centimeter. Margins entire in the The present species is not abundant in the ·cuneate basal half of the leaf. Above the mid Ripley collections. An1ong previously de dle the margins show seven or eight relatively scribed species it is close to the smaller forms 80 large crenate teeth, increasing gradually in of Oelastrophyllum crenatum Heer and 80 size upward and then diminishing toward the Oelast1·ophyllum nerwberryanu1n Hollick. rounded crenate tip. Petiole practically want Both of these are normally larger, and both ing, not extending beyond the decurrent mar have been recorded from beds at numerous gins. Midr~b straight, relatively stout below, horizons. Thus Oelastrophyllum crenatu1n is tlun above. Secondaries, seven or eight thin found in the Pa.toot beds of Greenland and in p~irs diverging frmn the midrib at acute angles the Raritan, Tuscaloosa, Magothy, and Black and can1ptodrome some distance within the Creek formations of the Atlantic Coastal Plain. margins. Tertiaries thin, mostly long fork Its smallest leaves are twice the size of the ing and inosculating veins subparallel with the present species and have n1ore numerous and secondaries, giving the leaves a .Llfyrsine-like less ascending secondaries. Oelastrophyllum character. newberryanum is also larger and relatively . The number of fossil s·pecies from the Ameri wider, with finer serrate teeth and decidedly can Cretaceous that have been referred to different tertiary venation. · Oelastrophyllum is very great. The genus Occurrence: MeN airy sand member of Rip makes its appearance in the latest deposits of ley formation, Perry place, Henry County; the Lower Cretaceous (Patapsco formation of Cooper pit, near Hollow Rock, Carroll County, Maryland and Virginia) and continues to be Tenn. prominent throughout the Upper Cretaceous, Celastrophyllum minimum Berry, n. sp. especially in the Atlantic Coastal Plain region. It. is much less common in western North Pla tP XII, figure 7 America and in Europe, being represented in Small, somewhat inequilateralleaves, lanceo Europe by forms usually referred to other late, falcate distad, widest in the basal half, genera, although Ettingshausen recorded two with an apiculate tip and inequilateral decur species of Oelastrophyllum from the Ceno rent base. Margins somewhat irregularly and manian of Niederschoena, in Saxony. Recently remotely serrate. Texture relatively coria l{ryshtofovich 79 has des•cribed a species from ceous. Length about 2.25 centimeters; maxi the Upper Cretaceous of Sakhalin Island, thus mum width about 5 millimeters. Petiole stout, hinting at a Mesozoic display of these forms in curved, about 3 millimeters in length. Midrib 79 Kryshtofovich,, A., Imp. Coli. Sci. Tokyo Jo•1r., vol. 40, so For a recent discussion of these species see Berry, E. W .. art. 8, p. 54, fig. 11, 1918. U. S. Geol. Survey Prof. Paper 112, pp. 108, 109, 1919. SAPINDALES 65 stout, not prominent, curved distad with a fold Occurrence: l\fcN airy sand member of Rip in the lamina, similar, for example, to many ley formation, Perry place, Henry County, existing ·species of Prwnus. Secondaries im Tenn. mersed, thin, numerous, camptodrmne. Celastrophyllum perryi Berry, n. sp. This small and characteristic form is only superficially suggestive of the associated Oelas Plate XII, figure 12 trophyllum variabilis Berry and seems to me clearly referable to the Celastraceae, although Broadly ovate leaves of considerable size, comparisons may be made with certain forms widest midway between the apex and base and of Ilicaceae. It is rare in the present collec about equally pointed at both ends, the base tions. decurrent to the short, stout petiole. Margins Occurrence : MeN airy sand member of Hip faintly and irregularly undulate. Leaf sub ley formation, Perry pl·ace, Henry County, stance thin. Length about 10 centimeters; Tenn.· maximum wi(J.th about 4.25 centimeters. Mid rib straight ~nd stout, especially in the lower Celastrophyllum cassinoides Berry, n. sp. half of the leaf. Secondaries thin, five or six Plate XII; figure 11 pairs, diverging from the midrib_ at angles of about 45° and camptodrome some distance Leaves oval-lanceolate, widest midway be within the margins. Tertiaries· forming open tween the apex and base and about equally arches as continuations of the secondaries and pointed at both ends. Margi~s conspicuously inside the margins, and open meshes within the undulate, leaf sub:stance thin. Midrib fairly secondaries; the secondaries are thin and im stout, curved, not prominent. Secondaries mersed, as is the fine meshed areolation. thin, remote, camptodrome a considerable dis In general facies this species is very similar tn nee from the margin ; those in the basal part to the associated Oelastrophyllum cassinoides of the leaf diverging at acute angles and as Berry, from which it differs in larger size, cending; those higher up diverging at angles more elliptical form, and more irregular and of 45° to 50°, straight about halfway to the less conspicuous undulate margins. The margins, where each forms a wide sweeping lower secondaries are less ascending, a feature arch to the secondary next above. Tertiaries correlated with the wider h'asal region. .The forming open arches around the margin and two are closely related, however, and it is pos an open mesh within the secondaries. Areola sible that they may represent the variants of tion very fine meshed. ·Length about 8 centi a single botanic species, a supposition that can meters; maximum width about 3.1 millimeters. be thoroughly tested only when very ~uch This well-marked species is not closely re more material is available for study. The lated to Inost of the Upper Cretaceous forms present form suggests the genus Elaeodendron referred to this form genus, many of which I Jacquin, which is confined to Africa in the ex have suggested should be transferred to. the isting :flor'a but which had a much more ex genus· Ternst?·oe'lnites, but as Oelastrophyllum tensive range in the late Cretaceous and Ter-: was proposed in the first instance for leaves of tiary. Celastraceae not generically determinable the -Occurrence: MeN airy sand member of Rip reference to it of these leaves of Ripley species ley formation, Perry place, Henry County, of undoubted Celastraceae is perfectly proper. Tenn. The present species is somewhat smaller and narrower than the associated Oelastrophyllum Celastrophyllum variabilis Berry, n. sp pe1"ryi Berry, with more pronounced undulate Plate XIII, figure 1 margins. A single dentate tooth is present on the left side of the figured spepimen, but this Leaves small, of somewhat variable form, is considered an anomaly. The species is not lanceolate, with pointed tip, prominently cre common in the collections, and the name is nate margins, and decurrent b'ase. Leaf sub- · give~ in allusion to the resemblance to certain stance of considerable consistency. Length existing species of Oassine Linne, a genus to 2.25 centimeters or less; maximum width, in which Elaeodendron Jacquin is often referred. the median :r;egion, 5 to 7 millimeters. Petiole 66 THE FLORA OF THE RIPLEY FORMATION short, stout, and curved. :M:idrib stout, curved. Family ACERACEAE Secondaries thin, camptodrome, sending short Genus ACER Linne branches to the marginal teeth. Tertiaries ob solete. Acer cretaceurn Berry, n. sp. These small somewhat protean leaves are Plate XIII, figure 2 considered to represent a Ripley species of Leaves of medium size, ovate, feebly trilo Celastraceae, although this assignment is not bate. Apex and lobes mucronate pointed. Base conclusive, for they may be compared with cordate. Leaf substance thin but of consider forms irt a variety of families. For example, able consistency. Length about 6 centimeters; the Ilicaceae contains si1nilar forms, as does maximum width about 6.75 centimeters. also the genus llfyrica. They also show a re ~fargins, except the entire cordate base, with semblance to some of the forms referred to unequal coarse dentate teeth, which become the genus Tenistroemites, and finally Vele mucronate where a primary or secondary vein novsky 81 has described a similar leaf from the runs to their apex. Petiole stout, curved and Chlomeke.r sandstone (Emscherian) of Bo expanded proximad, 2.65 centimeters in hemia as Rhus cretacea. length. Midrib stout, becoming thin distad, Occurrence: MeN airy sand member of Rip prominent on the lower surface of the leaf. ley formation, Perry place, I-Ienry County, Lateral primaries diverge from the midrib at Tenn .. angles of about 50° to 55° at its extreme Genus PACHYSTIMA Rafinesque~ base; they are stout and run to the tips of the lateral lobes. Secondaries frmn the mid Pachystirna? cretacea Berry rib, two or three subopposite pairs, diverg Pachystima? oretaoea Berry, U. S. Geol. Survey Prof. ing at angles of about 50°, stout at first, rap Paper 84, p. 49, pl. 10, fig. 6, 1914. idly thinning, craspedodrome. Secondaries from the lateral primaries, three on ·proximal Remains identical with the type material of side and one on distal side, rather straight, this species are present in western Tennessee. craspedodrome, terminating in enlarged mu It was described by me in 1914 as follows: cronate pointed teeth. Leaves of small size, oblong or obovate in outline, This is a very handsome and characteristic with a rounded apex ·and a narrowed descending base; species of rna pie, quite d~stinct from the ear about 2 centimeters in length by 3 millimeters in lier leaves that have been referred to tlv~ greatest width. Texture coriaceous. Petiole short genus Acer, all of which hitherto recorded and stout. Midrib stout. Secondaries numerous, fine, immersed, ascending. from the Upper Cretaceous are of doubtful identity. In so far as the present collections It may well be doubted whether these Up go this species is not common in the Ripley, per Cretaceous forms represent the modern but it should be· realized that negative evidenc;a genus Pachystirna, and it would probably have of this kind has very little real weight. The been better if they had been referred to. the maples are essentially Tertiary trees, although forin genus Oelastrophyllum. they evidently originated and enjoyed a con There are only two existing species of siderable dispersal prior to their known Pachystima, one from the East and one from Eocene occurrences. the Rocky Mountain area. They are shrubs of In addition to the present Ripley species dry situations, with evergreen, more or less eight other Cretaceous leaves have been re revolute leaves. As above stated, the generic ferred to Acer with more or less certainty. A identification is doubtful, but a fossil species review of these, shows that for most of of this genus is known from the Miocene of them the evidence of identity is far from con-· Colorado. elusive. Thus Acer paucidentatum Hollick 82 Occurrence : MeN airy sand member of and Acer min'utum Hollick,83 from the Upper Ripley formation, Perry p~ace, Henry County, Cretaceous of Staten Island, N. Y., and Cliff- Tenn. 8l1 Hollie!{, Arthur, New York Aead. Set. Trans., vol. 16, p. 81 Velenovsky, Josef, Die Flora der bohmischen Kreldefor 132, pl. 14, figs. 2, 3, 1897. · mation, pt. 4, p. 7, pl. 4, figs. 7-12, 1885. sa Idem, vol. 12, p. 35, pl. 3, fig. 6, 1892. RHAMNALES 67 wood, N. J·., are very doubtful and in my of this genus but conforms to the type, with judgment do .not represent Acer j Acer am numerous well-marked secondaries,' campto boyense Newberry,84 based on fruits from New drome close to the margins, and with numerous Jersey and Marthas Vineyard, probably rep subparallel percurrent tertiaries. It does not resents Pinus j Acer sa.skatchewense ·nawson,scs appear to have been a common form in the from Northwest Territory, is of uncertai.:.1 Ripley. value; and .Acmites 1n~~ltiformis Lesquereux,86 Occurrence : l\1cN airy sand n1ember of Rip from the Dakota sandstone, and Acer antiquum ley formation, Perry· place, Henry County, Ettingshausen,87 from the Cenomanian of Tenn. Saxony, are very doubtful, although · they Genus ZIZY:PHUS Linne somewhat resemble one another and show a certain resemblance to N egundo. Zizyphus laurifolius Berry The only Cretaceous species outside the Rip Plate XIII, figure 4 ley that are of unquestionable identity are Zizyphus lctu-rifolius Berry, U. S. Geol. Survey Prof. 88 89 Acer caudatum and Aoer edentatum. . Paper 84, p. 116, pl. 21,. fig. 7, 1914; Prof. Paper Occurrence : MeN airy sand member of Rip 112, p. 113, 1919. ley formation, Cooper pit, near Hollow Rock, Leaves linear-J.anceolate, with entire mar Carroll County, Tenn. gins, about equally and abruptly pointed at the Order RHAMNALES apex and base. Length 9 to 11 centimeters; maximum width 1.5 to 2 centimeters. Petiole Family RHAMNACEAE missing in all specimens so ·far discovered. Genus RHAMNUS Linne Primaries three, eli verging at or ·a short dis tance above the base. Midrib stouter and much Rhamnus ripleyensis Berry, n. sp. more promin~nt than the lateral primaries, Plate XIII, figure 3_ which diverge from it at acute angles and sweep upward, pa.rallel with the lateral mar Leaves lanceolate, somewhat irregular in gins, to or above the middle of the leaf, where form, with a narrowly pointed tip and a more they join short branches from the lowest sec abruptly pointed base. Margins entire but ondaries. Secondaries thin; three or four al irregularly rounded. Texture coriaceous. ternate to opposite pai1~s diverge from the mid Length about 9 centimeters; 1naximum width, rib at acute angles in the upPer part of the slightly below the middle, about 2.2 centime leaf and sweep upwa.rd in long, regularly and ters. Petiole short and stout. Midrib very but slightly arched camptodrome curves. A stout and prominent on the lower surface of few thin camptodrome secondaries diverge· at tl13 lea£, slightly flexuous. Secondaries ;nu acute angles from the outer sides of the lateral merous, stout, prominent, subparallel, diverg primaries. Tertiary venation obso~ete. Tex ing from the _midrib at angles of about 65 °, ture coriaceous. curvillg regularly, camptodrome close to the This species was described from the lower m.argins. Tertiaries thin, very close-set, per Eutaw of the Chattahoochee drainage basin, current. ''here it was represented by a scant amount of This species is distinct among the various rather poorly preserved material. The ma Upper Cretaceous species that have been re terial from western Tennessee is similra.r ly ferred to Rhamn~l·S. It is relatively more nar scanty but somewhat better preserved and en rowly elongate than most of the other species ables me to amplify somewhat the previous characterization of the species. It is easily B4 Newberry, J. S., U. S. Geol. Survey Mon. 26, p. 106, pl. 46, figs. 5-8,. 1896. distinguished from the associated Ripley spe S& Dawson, J. W., Roy. Soc. Canada Trans., vol..3, sec. 4, p . cie$ by its entire margins and may be readily . 16, 1886. so Lesquereux, Leo, U. S. Geol. Survey Mon. 17, p. 156, distinguished from all the other Upper Creta pl. 34, figs. 1-9, 1892. ceous entire-margined 5pecies except Zizyphus s1 Ettingshausen, Constantin von, Kreidefl.ora von Nieder 90 schoena, p. 2-.3, pl. 3. fig. 17, 1867. cliff~ooodensis Berry by its· size and elonga.- ljlj Heer, Oswald, }j"lora fossilis arctica, vol. 7, p. 38, pl. 65, figs. 1, 2, 1883. oo Johns Hopkins Univ. Circ., new ser., No.. 7, p. 88, fig. 5, Bll Idem, p. 39. pl. 65, fig. 3. 1!)07. 68 THE FLORA OF THE RIPLEY FORMATION tion. Zizyphus cliffwoodensis, which comes sence from the numerous Cretaceous plant from the.Magothy formation of New Jersey, is bearing beds of Europe is a fact of consider still larger, ovate-lanceolate, and with numer able significance. No less than ten diff·erent ous conspicuous percurrent tertiaries. species have been recorded from the American Occurrence: McNairy sand member of Rip Cretaceous, and although some of these are of ley formation, Cooper pit, llh miles south of doubtful validity, as, for example, Zizyphus 92 Hollow Rock, Carroll County, Tenn. oblongus Hollick, the majority are well char acterized and undoubtedly related to Zizy Zizyphus ripleyensis Berry, n. sp. phus. The only Eurasian Zizyphus known from the Cretaceous is the forn1 identified as Plate XIII, figures 5-9 Zizyphus dakotensis Lesquereux by ICryshto Leaves of variable size, l'anceolate or ovate fovich, 93 from the Cenomanian of the Ural lanceolate, widest slightly below the middl~, region, and the indistinctness of his figure pre tapering downward to the acute base, which is cludes passing a critical judgment on the iden sometimes inequilateral, and upward to the tification, although it appears to he correct. slender, acute, and slightly extended tip. If it is correct, then the oldest known form is Length 4.5 to 9 centimeters; maximum width from the Oenomani.an of the Ural region, on 1.5 to 2.4 centimeters. Margins entire for _the assumption that the age of the beds in their basal third, above which they 'are beset which it was found was correctly determined, with somewhat remote serrate teeth, which are which is doubtful, for all the American oc often unequally spaced and vary in promi currences are post-Raritan and pre~uma.bly nence not only from specimen to specimen post-Cenomanian. Although Zizyphus is re but in a single individual. Primaries three, corded from the Cretaceous of l(ansas and diverging at or near the base, the midrib being New Mexico and doubtfully from Alaska, it much the stoutest. The lateral primaries di makes its greatest display both in numbers verge at .'acute angles, curving upward parallel and variety ~n the Atlantic Coastal Plain. with the lateral margins and becoming thinner With the dawn of the Eocene the genus ap in the distal third of the leaf as they arch peared in force in Europe, although equally from end to end of the straight outer branches abundant on both continents, ten species being of the superjacent secondaries. Secondaries recorded from each, the bulk in both being in from the midrib two or three pairs in the the lower Eocene. upper part of the leaf, diverging at acute There are 10 Oligocene and 20 Miocene angles 'and sweeping upward in long ascending species, the genus being clearly cosmopolitan curves. Tertiaries thin, approximately trans in the J\1iocene epoch. The Pliocene records verse, either percurrent or inosculating mid shrink to 4 or 5 species, all European, although way between lateral primaries and the midrib ; the presence of fruits in the Pleistocene of those from the outer side of the lateral pri New Jersey and Florida would seem to indi maries form abruptly camptodrome arches and cate continuous residence in North America send off short branches to the margin'al teeth. until comparatively recent time. The modern Texture coriaceous. species number two score, largely shrubs, many This well-marked species is readily distin of which are prostrate or scrambling, and a guishable from the other known Cre.taceous few small trees, with their maxim~m develop species, the only one of which that at all re ment in the Indo-l\1alayan region,' alt.hough sembles it being Zizyphus dakotensis Lesque present in the Tropics of eastern Asia, Amer reux,91 a prevailingly larger and much more ica, Africa, and Australia, as well as in Medi elongated fonn with thinner primaries and terranean Europe and South Africa. more prominent teeth. Among recent species Some of the existing species are extensively the leaves of Zizypl},~ts vulgaris Lamarck are· planted, and Zizyphus vulgaris has become not unlike the present fossil species. The great naturalized in some of our Gulf States, so that v~riety of leaves of Zizyphus in the Upper the extinction of the genus in North America Cretaceous of North America and their ab- ~3 Hollick, Arthur, U. S. Geol. Survey Mon.. 50,. p·. 92, pl. 34, figs.. 9, 10, 1906. 01 Lesquereux, Leo, The flora of the Dakota. group: U. S. oa Kryshtofovich,, A., Acad. imp. sci. St.-Petersbourg Bull., Geol. Survey Mon. 17, p. 167, pi. 36, figs. 4-7, 189·2. 6th ser., p. 608, fig. 7, 1914. RHAMNALES 69 .ilJ late Pleistocene or postglacial time is an from the base of the midrib at acute angles, other one of th~ unsolved mysteries of plant ascending rather straight about halfway to the h]story. The accompanying sketch map (fig. tip, where they join an outward branch from 6) shows the limits of natural range of Zizy the lowest secondary. Secondaries four or 'phus in the existing flora. It will be noted five subopposite remotely spaced pairs, stout., that the Cretaceous occurrences are mainly camptodrome; the lowest pair longest and occidental, and from early Eocene time on regularly curved; the others straight halfway ward the genus is cosmopolitan in the North to the margin and then each with a single flat ern Hemisphere. The late Tertiary occur camptodrome curve to the secondary next rences are largely European and would seem above. Secondaries from the outside of the to indicate the waning of the genus in the primaries forming r:ather straight-sided and 'VQStern. Hemisphere, although the presence of flat-arched areas. Tertiaries relatively ·Stout fruit remains in the late Pleistocene of New and prominent, forming arches immediately Jersey and Florida indicates straggling species within the margins and an open, prevailingly down to almost recent time. quadrangular mesh within the secondaries and FIGURE 6.-Sketch map of the world summarizing the geologic history of Zizyphus. Area within heavy line, maximum existing development; area within th.ln lines,, limit of existing natural range; northeast-southwest diagonal lines, regions of Cretaceous occurrences ; northwest-southeast diagonal lines, regions of Tertiary occurrences ; solid circles, areas of .Pleistocene occurrences Occurrence : J\1cN airy sand member of primaries. Length· about · 10 centimeters; Ripley formation, Perry place, Henry County, maximum width about 4.5 centimeters. Tenn. (common) ; Cooper pit, near Hollow This characteristic form is named in ap Rock, Carroll County, Tenn. (rare) . preciation of the Perry family, on whose prop erty the fossiliferous Ripley clays .are exposed Zizyphus perryi Berry, n. sp. and whose intelligent interest was most help Plate XIII, figure 10 ful .in making the collections. It is a larger Leaves fairly large, ovate, widest below the and relatively wider species than Zizyphus middle, tapering in almost straight converg ripleyensis Berry, with much smaller mar ing lines to the acute tip; the lower lateral ginal teeth. The remains· are usually broken margins full and rounded, and the base broad but are not uncommon in the collections. The ly cuneate. J\1argins irregularly dentate to specimen figured is the only one showing the serrate, the teeth tending to be most promi whole leaf, one other upper :P,alf and four nent below the middle, in general with shal basal halves of leaves representing the rest of lowly rounded, almost equilateral sinuses. the material referred to this specie~. Texture subcoriaceous. Petiole missing in all Occurrence: McNairy sand 1nember of Rip the specimens. J\1idrib stout, prominent.~ ley formation, Perry place, Henry County, curved. Lateral primaries stout, diverging n Tenn. 100913-25--6 70 TI-fE "FLORA OF THE RIPLEY FORMATIO:X Family VITACEAE tal side, entire on proximal side in the· small Genus CIS SITES Heer leaf figured, in the other with· a subordinate Cissites crisp us V elenovsky lobe developed to a greater or less degree; basal lobes narrower and less extended than the Cis.sites cr·~sznts Velenovsky, Die Flora cler bohm ischen Kreideformation, pt. 4, p. 12, pl. 4, others, obtusely rounded, .at right angles to fig. 6, 1885 (not Newberry, 1896, nor Berry, midrib, without subordinate lobes, although the 1911). Berry, Torrey Bot. Club Bull., vol. 43, beginning of a lobule is indicated on the right p. 296, 1916. proximal side in the 1nedimn-sized specimen Oissites crispus was identified fr01n the figured. Length from 5.5 to 12 centin1eters: Raritan formation of New Jersey by New maxin1um width 5 to 11.5 centi1neters. Petiole. berry and from the Magothy formation of missing in Tennessee material, long and stout that State by Berry, but neither of these oc in the type, at least 3 centimeters in length. currences represents the European form·, so \: enation quinquepa.lmate from point at or that recently I have made the~ the basis of near the base. JV[idvein fairly stout and nea:rly a new species, Oissites newberryi. 94 straight. ~1:iclveins of main ·lateral lobes eli A perfectly well characterized,. small-leafecl verging from central vein at angles of 25° to form that appears to be identical with the 40°, nearly straight or somewhat curved, as Bohemian type -of Oissites crispus is, how shown in specjmens figured; midnb of basal .ever, present in the Ripley formation of west lobes of lesser caliber, tending to be inserted ern Tennessee. It differs from Oissites new just above the base of-the lateral primaries, di berryi in its relatively shorter and wider form, verging at angles· of 30° to 40° and curving its crenate instead of serrate or dentate teeth, o11tward; iri the single type specimen the basal its les~ ascending secondaries, and its cordate lobes are somewhat wider than in the Tennes base. see material and their midvein is nearest the The type and only other known occurrence distal margin, where as in the Tennessee speci of Oissites crispus is in the Chlomeker beds mens they are nlore nearly centra.! in position. (Emscherian) of Bohemia. Secondaries rather remote and slender, diverg Occurrence: ~IcN airy sand membe.r of Rip ing at acute angles and curving upward m ley formation, 21;2 miles southwest of Selmer, gentle long curves; those that run to the tips MeN airy County, Tenn.. of subordinate lobes somewhat stouter and Cissites panduratus Knowlton longer; the rest camptoclrome except those that run to the sinuses, which unite to form a more Plate XIV or less well-developed margina.I hem, such as Cissites pandttratus Knowlton, U. S. Geol. Survey is so well developed in the n1odern and some of Prof. Paper 101, p. 27 4, _pl. 49, fig. 10, 1918. the fossil lobate leaves of Sa8safras. The Leaves of variable size, multiloba.te, prevail tertiary areolation is not discernible. The tex ingly quinquelobate, derived from a trilobate ture is thin and not coriaceous but scarcely ancestor. by the development of subordinate merits .the term membranaceous applied to the basal lobes on either side, which in the Tennes type. see material tend to be shorter than in the type This species was based t1pon the single speci material from the _Raton Mesa region, although men .figured by Knowlton, from the Vermejo the medium-sized specimen from Tennessee fig formation of southeastern Colorado. It is ured here is almost an exa.ct duplicate of the very abundant in the MeN airy sand member type. Main sinuses extending more than half of the Ripley formation of Tennessee, so that way to the base, broadly and rather evenly ·the description can be somewhat amplified and rounded; middle lobe obovate, with a lateral the range in size and outline c'an be evaluated. lobe of greater or less development on either Knowlton compared the species with Acer, ·side, obtusely pointed; main lateral lobes un Liriodendron, Oissus, and V itis and pointed symmetrically obovate, narrower than the ter out its resemblance to Oissite?J formosus Heer, minal lobe, obtusely pointed, entire on the dis- more especially to the trilobate forms from the Dakota sandstone ascribed to that species by 04. Berry, E. W., Maryland Geol. Survey, Up·per Cretaceous, p. 856, 1916. Lesquereux. Heer's type of Oissites formosus, D RHAMNALES 71 as well as the forms from the Raritan forma ancestral decurrent basal margins and tripal tion of New Jersey referred to this species by mate suprabasilar primaries is Oissites for Newberry are distinctly not trilobate 'and, nwsus 1nagothiensis Berry, from the somewhat though they show considerable variation, are later J\1agothy formation of Maryland. Ois· extremely close to the present species,. as can sites panduratus, from the approximately con be appreciat.ed by referring to the restorations temporaneous but later horizon in Colorado· published in 1917. 011 There is not the slightest and western Tennessee, may be derived from doubt that Oissites pa.nduratus is properly .re Heer's type by the truncation of• the base, the ferred to Oissites, and I consider it specifically deepening of the sinuses, and the correspond distinct from Oissites formosus, which has de ing .downward and outer movement of the parted less from the original trilobate stock; lobule of the lateral lobes, which become dis the b'asal lobes are still lobules of the lateral tinct though somewhat subordinate lobes with lobes, from which .they are separated by a shal a corresponding change in position of their low sinus and are not yet independent basal midveins. We have here as clearly marked lobes; the basal margin is decurrent and not an evolutionary series as can be cited from truncate; and the primaries are three and dis allied fields of paleozoology, for although we tinctly suprabasil'ar. are dealing with only foliar. remains they pre~ An interesting series of variations from this serve their generic facies through rather wide central and ancestral stock of the Atane beds variations, and although a.t .each recorded of 'vestern Greenland has since come to light. horizon they are variable in form, these varia In the Raritan formation of New Jersey it is tions culminate in successively· younger occur represented by forms distinctly more sublobate, rences, so that the latest form is most decidedly and the same tendency resl.llts in the form distinct from the earliest. It would perhaps Oissites dentatolobat~ts, from the Dakota sand be well to consider the Raritan form as a con stone, described' by Lesquereux, which is not stant enough variant of Oissites formosus to unlike a form from the Perucer beds ( Ceno merit a distinct varietal name, but that is of manian) of Bohemia described by Velenovsky. little consequence in the present connection, as A narrow, less sublobate, and more elongate I wish to emphasize here the place of origin variant of Oissite8 formosus that preserves the and subsequent modification and radiation of this series of forms. This might be repeated 05 Berry, E. W., Maryland Geol. Survey, Upper Cretaceous, pl. 79, figs. 1, 2, 1917. diagrammatically as follows: Cissites formosus Heer (Atane beds of Greenland) ----\ \ \ ---- \ \ \ \ Variant in ;r{a'~ ----formation of \ \ \ New Jersey . \ \ Cissites dentatolobatus Lesque- . reux (Dakota sandstone of \ \ Kansas) Cissites formosus magothiensis \ Berry (Magothy formation of New Jersey) . Cissites panduratus Knowlton . (Vermejo formation of Colo rado and Ripley formation of 'fennessee) The diagram clearly shows .the relationship obvious that none of. these specific regions cor of these forms, but it; implies an Arctic origin, respond with the original center of radiation, which does not fully accord with the facts as I and if the current conceptions of mid-Creta understand them. As allied species occur in ceous paleogeography are considered it will be the Atane beds of Greenland, the Cenomanian seen that the region which exactly fits the re of central Europe, and beds of approximately quirements is the central or north Asiatic re the same age in New Jersey and J(ansas, it is gion, from which it spread southwestward into 72 THE FLORA OF THE RIPLEY· FORMATION Europe, across the Arctic mainland to Green-~ The diagram may be recast more correctly and land, and across the Bering region to Kansas. somewhat amplified as follows: Hypothetical Asiatic ancestor ------~~--- \ \ Cissites form~eer (Atane \ \ / beds of Greenland) ~ . \ \ \ Raritan variant\ \ \ \ (New Jersey) \ \ \ \Cissites vitifolia Velenovsky (Ceno- Cissites formosus mago- manian of Bohemiaj thiensis (Mag7thy, \ · Maryland) Gissites dentatolobatus _ Lesquereux (Dakota sandstone of Kansas) Co1orado Tennessee Cissites panduratus Knowlton Occurrence: McNairy sand member of Rip teeth. Tertiaries thin but well marked, ley formation, Perry place, Henry County, curved, percurrent or inosculating in the Tenn. region midway between secondaries nnd pri Order MAL VALES maries. Areolation fine meshed, largely. im Family TILIACEAE mersed in the leaf substance. This handsome species is obviously new. It Genus GREWIOPSIS Saporta is a type of leaf often referr~d to the genus Grewiopsis ripleyensis Berry, n. sp. Populus, and similar forms have sometimes been considered as representing nonlobate Plate XV, figure 2 leaves of Platanus, but it is certainly not re Ovate to narrowly cordate leaves of medium lated to these two genera and is obviously re large size, with pointed tip and a somewhat lated to a variety of forms in· the order Mal cordate base. Substance thin but firm and vales. In order not to be too specific in its polished. Margins with large, somewhat irreg generic determination it is referred to the form ular serrate teeth. Length about 10 centi genus Grewiopsis, which was instituted by meters; ma~imum width, near the base, about Saporta in 1866 for leaves referable tn the 7 . centimeters. Petiole missing, obviously order Malvales but riot positively referable to stout. Midrib stout, fairly prominent. Lat existing genera. He compared these leaves, eral primaries three on each side, diverging at which were abundant in the " Pal~ocene " of acute angles from the top of the petiole. Out Sezanne, France, with the existing genera side pair short and straight, directed at right Sida, A.butilon, Pterospermum, Dombeya, angles to midrib, craspedodrome. Second pair Grewia, Sparmannia, Luhea, and Tilia of this also short and straight, about 50 per cent order. They have also been compared, on what longer than the outside pair, terminating in seem to me doubtful grounds, with the extinct marginal teeth in the lower half of the leaf. genus Oredneria Zenker. About a score ·of Third pair longer, diverging from midrib at species are known, largely from the early angles of about 35°, curving upward and ter Eocene of Europe and America. Two species Ir..inating in marginal teeth in the upper half are known from the Dakota.sandstone, Grewi of the leaf, sending off on the outside regu opsis aequidenta-ta and Grew·iopsis flabellata. larly· spaced craspedodrome branches. Sec both described by Lesquereux. Lesquereux de ondaries from the midrib generally three alter scribed a third species, Grerwiopsis mudgei~ r..ate pairs, the lowest 3 centimeters or more but this was an unintentional redescription· of above the lateral primaries, diverging from the typ~ of Grewiopsis aequidentata. G~rewi the midrib at angles of about 45° or 50°, curv opsis flabellata has been recorded from the ing regularly upward and craspedodrome, Magothy formation in the Coastal Plain of sending outward thin branches to the marginal Maryland. A third species, Grewiopsis cle- MALVALES 73 bur'ni, has been recorded by Lesquereux from The species is obviously distinct from pre the Cretaceous of the Montana group, and viously described forms and readily recog there are ·two well-characterized species in the nized. The only fossil leaf to which I have Tuscaloosa formation of Alabama. noted any resemblance is a form from the lower The two Tuscaloosa forms, as we.ll as the Eocene which Ward 96 identified with Acer present species, are more closely allied to cer ·indivisum Weber, an .obvious J;Distake. This tain modern. genera of the family Tiliaceae. has since been considered an abnormal form These .are G1·ewia Linne, which has about 90 of the American leaves that have been referred species, mostly tropical, in the Old World; to Platanus guillelmae Goeppert. . Hockenya 'Villdenow, which has about 15 The present species finds its nearest living species in the American Tropics, extending analogues in the tiliaceous genera Grewia, T?'i from the Antilles and Central America to umfetta, ·and H ocke.nya. Brazil; and Trium}etta Linne, with about 60 Occurrence: MeN airy sand member of Rip species in the Tropics of both hemispheres~ ex·· ley formation, Perry place, Henry County, and tending northward in America to the Antilles. Cooper pit, near Hollow Rock, Carroll County, Occurrence: l\icN airy sand member of Tenn. Ripley formation, Cooper pit, near Hollm:v Rock,. Carroll County, Tenn. Family STERCULIACEAE Grewiopsis inequilateralis Berry, n. sp. Genus STERCULIA Linne Plate XV, figure 1 Sterculia snowii tennesseensis Berry Leaves smaller than those of Grewiopsi:s Sterculia snowii tennesseensis Berry, Torrey Bot. Club riJJleyensis, markedly inequilateral in all the Bull., vol. 43, p. 296, pl. 16, fig. 5, 1916; U. S. four specimens contained in the present col Geol. Survey Prof. Paper 112, p. 117, pl. 32, fig. lection, oval, with a bluntly pointed, not ex 1, 1919. tended tip and an obliquely truncated rounded Leaf bilobate, with a bluntly pointed base base. Texture much more coriaceous than in and gradually narrowed acuminate recurved the preceding species. Length from 3.5 to 5.5 apical lobes. . Length about 11 centimeters; centimeters; width in the median region 2.5 width of the entire basal part of the leaf 2.5 to 4 centimeters. Margins entire below; above to 2.75 centimeters; width of lobes 1.1 to 1.6 the truncated base with large dentate teeth, centimeters. Margins entire. Texture sub showing a tendency to become slightly sub coriaceous. Sinus extending· halfway to the lobate with rounded shallow sinuses. Petiole base or less, open, na.rrowly rounded. Midrib long and stout. l\1idrib stout and prominent. Etout, flexuous. Lateral primary stout, diverg Lateral primaries fron1 the top of the petiole, ing from the midrib at an acute angle about 3 usually two on one side and one on the other, centimeters above its base. Secondaries thin, diverging frmn one another at acute angles. largely immersed, diverging from the prima The odd one is short and not much curved, ries at wide angles and at regular intervals, craspedodrome. The pair a.re curved, ascend arching in a camptodrome manner near the ing, and craspedodrome; all send regular margins. curved secondaries on the outside to the mar This striking form is unfortunately repre· ginal teeth. Beyond an interval above the sented by only two specimens, both of which primaries the midrib sends off a pair of curvt!d are bilobate, although, like so many other fossil craspedodrome secondaries, ancl there are two and existing species of Sterculia, it may well additional small pairs in the upper part of the have· varied froni entire to trilobate. Among leaf. Tertiaries percurrent and not promi previously described fossil forms it may be nent: areolation obsolete. compared with the species from the Magothy The largest specimen of this species is fig formation in New Jersey and Maryland, Ster ured, and it should be noted that the peculiar culia minima Berry,· which is a smaller and base is normal and not broken in any way. more variable form, or with tlie species from One of the smaller specimens is less iriequilat the Dakota sandstone, Sterculia mucronata eral, but the other two are similar in thi"s re 00 ~pect to the figured specimens. Ward, L. F., U. S. Geol. Survey Bull. 37, p. 66, pl. 29, fig. 5, 1887. 74 THE FLORA OF THE RIPLEY FORMATION Lesquereux and Sterculia· snowii Lesquereux. tip and a cuneate base decurrent on the winged Sterculia snowii, though often much larger petiole. Length from 5 to 18 centimeters; and with five lobes, is extremely variable. Two maximum width, below the middle, from 1 to named varieties of it have: already been recog~ 3 centimeters. Leaf substance thin. Margins nized, and the general character and venation regularly crenate except for a short distance of this Tennessee for.t;n leads me to conclude proximad, where ·they are entire~ Teeth vari that it represents another variety of this pro~ able but uniform on single leaves, regularly tean species. spaced, either close set or ·remote. Petiole long, Occurrence: MeNairy sand member of Rip~ conspicuously winged, 2 centimeters long in a ley formation, 2! miles southwest of Selmer, medium-seized specimen. Midrib s.tout, promi MeNairy County, Tenn. nent. Secondaries thin, diverging from the midrib at· angles of about 70° at regular in Order PARIETALES tervals and abruptly camptodrome in the mar Family DILLENIACEAE ginal region, sending short branches to the Genus DILLENITE~ Berry .marginal teeth. Tertiaries obsolete. This is one of the commonest species in the Dillenites cretaceus Berry, n. sp. Ripley, being represented by a large amount of Plate XV, ~gure 5 n1aterial. Superficially it is much like the form Leaves of medium size, ovate, widest below known as Oelastrophyllum carolinensis, but it the middle and tapering upward to the acumi differs in the elongated winged petiole, the ex nate tip. Base rounded or broadly cuneate. tended tip, and the less numerous and less as Margins dent'ate. Length about 8 to 9 centi~ cending secondaries. It appears to be ances meters; maximum width 3.4 to 4 centimeters. tral to the lower Eocene species Ternstroemites Petiole medium stout, about 1 centimeter in preclaibornensis Berry, of the Wilcox group, length. Midrib stout and prominent. Sec which is in turn ancestral to Ternstroemite.~ ondaries thin but prominent, numerous, sub claibornensis Berry, of the Claiborne group. opposite to alternate, subparallel, campto Occurrence: MeN airy sand member of Rip drome; about nine to twelve pairs diverge ley formation, Perry place, Henry County, from the midrib at angles of about 45°. Ter Tenn. tiaries thin, percurrent. These leaves are clearly allied to the several Temstroemites tennesseensis Berry, n. sp. species of Dille11ites from the lower Eocene of Piate XVI, figure 3 the Mississippi emb'ayment which I have de scribed and which represent early forms of Leaves of medi urn size, e'Ili ptical-lanceolate, Dilleniaceae. The present species is the only with a bluntly pointed apex and a similarly one thus far recognized in the Upper Creta pointed cuneate base. Length 8 centimeters; ceous. It differs from the Eocene species in maximum width, halfway between the ·apex having camptodrome secondaries and in send and the base, 3.7 centimeters. Leaf substance ing short branches into the marginal teeth, but thin. Margins regularly and somewhat . this feature is partly seen in Dillenites micro coarsely crenate except proximad, where they .dentatus, the type of the genus. · The present are entire for about one-fourth their extent. form is not uncommon in the Ripley. Midrib stout, straight, and prominent on the Occurrence: MeN airy sand member of Rip· lower surface. Secondaries thin; about 12 ley formation, Cooper pit, near Hollow Rock, pairs diverge from the midrib at wide angles, Carroll County, Tenn. as n1uch as 60° in the median part of the leaf; they are approximately straight ·and parallel Family TERNSTROEMIACEAE . at regular ·intervals and are joined by a Genus TERNSTROEJd:ITES Berry camptodrome loop only about two-thirds of Temstroemites ripleyensi's Berry, n. sp. the distance to the margin. Tertiaries obsolete. This species is typical of the group of leaves Plate XV, figures 3, 4; Plate XVI, :figur.e 1 which are so common in the Upper Cretaceous Leaves of variable size, lanceolate, ·with an of the Atlantic Coastal Plain and which have extended and gradually narrowed acuminate heretofore been referred to Oelastrophyllum~ • Til Y:\IELEALES 75 ::1,lthough they are obviously not .Celastrarceae lower Eocene T ernstromnites eoligniticus but appear to represent the Ternstroemiaceae. Berry, of the vVilcox group of the embayment The present species shows points of resem region. blance to Oelas.trophyllum crena.tum Heer,97 Occurrence : MeN airy · sand 1nember of 0. 'l.tnclulatum Ne·wberry, 9 '~ and 0. g1·andijoliurn Ripley formation, Perry place, Henry County, Newberry,09 without being identical with either. and Cooper pit, near IIollow Rock, Carroll They range in age from the Tuscaloosa and County, Tenn. · Uaritan formations upward and are congeneric, Order THYMELEALES showing marked contrasts with the other Coastal Plain species that have been referred Family LA URACEAE to Oelast1'0JJ1hyll7.t/m. Genus CINNAMOMUM Blume The type of" Te1·nstroemites represented by Cinnamomum newberryi Berry the foregoing species appears to have been waning during Ripley time, and the contrasted Plate XVI, figure '5 types represented by Tern8troe1nites ripleyensis Oinnamomu·m newbe1-ryi Berry, Torrey Bot. Club and T. m·etaceus, equally old, appear to have Bull., vol. 38, p. 423, 1911; New Jersey Geol. furnished the stock of the lower and middle Survey Dull. 3, p. 150, pl. 16, _fig. 3, 1911; Eocene species in the embayment region. U. S. Geol. Snrvey Prof. Paper 84, 'pp. 54, 117, pl. 9, figs. 12, 13 ; pl. 21, tigs. 9-11, 1914; Occurrence: MrcN airy sand membei· of Rip Maryland Geol. Survey, Upper Cretaceous, p. ley formation, Cooper pit, near Hollow Rock, 860, pl. 71, fig. 6, 1916 ; U. S. Geol. Survey Carroll County, Tenn. Prof. Paper 112, p. 118, pl. 21, figs. 6-9, 1919. Oinnanwmum sezannense Beer, Flora fossilis arctica, Ternstroemites cretaceus Berry, n. sp. vol. 6, Abt. 2, p. 77, pl. 19, fig. 8; pl. .33, figs. 11, 12, 1882 [not \Vatelet] ; idem, vol. 7, p. 30 Plate XVI, figure 2 'pl. 61, fig. la, 1883 [not Watelet]. · Lesquereux, The flora of the Dakota group: Leaves of variable size, broadly lanceolate, U. S. Geol. Survey Mon. 17, p. 107, pl. 12, with an acuminate tip and a narrowly cuneate fig. 7, 1892 [not fig. 6, which is a leaf of base. Margins regularly but remotely serrate. G-innamomum. mentbranaceum, (Lesquereux) Texture subcoriaceous. Length from 10 to Hollie!~]. Dawson, Roy. Soc. Canada Trans., 1st ser., vol. 2, 17 centimeters; maximum width, in middle sec. 4, p. 64, pl. 13, 1ig. 58, 1894 [not ·watelet]. region, 1.5 to 3.25 centimeters. Petiole miss Hollick, Torrey Bot. Club Bull., vol. 21, p. 53, pl. ing. :.Midrib stout, straight, prominent on tlu~ 180, figs. 5, 7, 1894 "[not \Vatelet]. lower surface of the leaf. Secondaries thin, Penhallow, Roy. Soc. Canada Trans., 2d ser., vol. close-set, straight, and subparallel, diverging 8, sec. 4, p. 46, 1902 [not Watelet]. Hollick, New York State Mus. Fifty-fifth Ann. from the midrib at angles of about ·40°, in Rept. for 1901; .p. r50, 1903. osculating or joined by percurrent tertiaries, Oinnamom1t,m intermedium Smith, On the geology of their tips joined by abn~pt and flat campt(• the Coastal Plain of Alabama, p. 348, 1894 drome loops, from which short. branches arc:! [nomen nudum]. sent to the teeth. Newberry, The flora of the Amboy clays: U. S. Geol. Survey Mon. 26, p. 89, pl. 29, figs. 1-8, 10, This well-marked species is readily . di~· 1896 [not EttingshausenJ. tinguished from the other Ripley· species o:t Berry, New Jersey State Geo~ogist Ann. Rept. for Ternstroe1nites. It appears to be descended 1905, p. 139, pl. 20, figs. 2-6, Hl06 [not Ettings from Oelast1•ophyllu-1n demtrrens Lesquereux, hausen] ; Torrey Bot. Club Bull., vol. 33, p. 179, of the Dakota sandstone and the Raritan and pl. 7, figs. 3, 4, 1906 [not Ettingshausen]. Hollick, 'The Cretaceous flora of southern New Tuscaloosa formations, and also to be related York and New England: U. S. Geol. Survey to Myrica hollicki Ward, of the Raritan for Mon. 50, p. 74, pl. 29, fig. 7; pl. 30, figs. 1, 2, mation, both of which are forms of Tern 1907 [not Etting-shausen]. stroe'lnites and not related to Oelastrophyllum Berry, Torrey Bot. Club Bull., vol. 37, p. 27, 1910 or Llfyrica. · It is ,probably ancestral to the [not Ettingshausen]. Leaves subcoriaceous, lanceolate to ovate 07 Heer, Oswald, Flora fossilis arctica, vol. 7,, p. 41, pl. 62, fig. 21, 1883. . lanceolate, varying greatly in size and conse os Newberry, J. S., U. S. Geol. Sur-vey Mon. 26, p. 102, pl. quently in appearance. Apex short, pointed, or 38, figs. 1-3, 1896. . 00 Idem, p. 104, pl. 19, flg. 8 ; pl. 21 figs. 1-4. more or less na1~row ly extended. Base broad, • 76 . THE FLORA OF THE RIPLEY FORMATION narrowed to the petiole. Primaries three, usu resent a single botanic species.· I regard this ally suprabasilar. This species is primarily statement as se.If-evident, ·and yet I can dis distinguished from Oinnamomum heerii cern no criteria except stratigraphic for spe Lesquereux by its ·relatively narrower form cific segregation. Olnnam01nu1n leaves are no and acute base. The Georgia material is not toriously variable both in the existing and in scanty but is for the most part extremely extinct floras, and the sa.me unsurmountable fragmentary. The single complete leaf is of difficulties are presented by a number of rerti the short type like that shown in Newberry's ary species, so called.. I have studied a great figure 10/ or like the leaves· from the· Arctic many specimens of Ci-nnamomum newberryi and from the Dakota sandstone which have f'rom earlier horizons and can not see in the usually been referred to Oinna1n01num sezan Ripley leaves which I refer to this species any nense W atelet. That they are not in the slight features that will differentiate them from the est degree related to this European basal abundant remains in the much older Raritan Eocene type may be readily seen by a compari and Tuscaloosa formations. In themselves son of the Amm~ican Cretaceous material with they are characteristic enough, as the. accom the figures of the true Oinnamomu1n sezan panying figure shows, and are not readily con nense as given by W atelet, Saporta, and fused with other CretaceOtts or with Eocene others.2 species of Oinnamomu1n. The present species, as he.re revised accord Occurrence: MeNairy sand member of Rip ing to the foregoing citations, has a remark ley formation, 21f2 miles south of Mifflin, Ches able range in the Upper Cretaceous. The ter County, and a quarter of a mile west. of Raritan formation of New Jersey is the oldest Buena Vista, Carroll County, Tenn. formation in which it has been found. Above the Raritan horizon it occurs in the Atane and Cinnamomum newberryi lanceolatum Berry, n. var. Patoot beds of Greenland, in the Magothy Plate XVI, figure 4 formation from Long Island to Mary land, in the Black Creek formation of North Carolina, Small slender leaves, lanceolate, widest be in the Middendorf arkose member of the Black low the middle and tapering upward to the Creek of South Carolina, in the Tuscaloosa extended acuminate tip. Base nar:row, slightly formation of Alabama, where it is especia1ly decurrent. Texture subcoriaceous. Margins abundant at certain localities, and in the Da strictly entire. Length about 4 centimeters; kota sandstone of l{ansas.. It appears to be maximum width about 8 millimeters. Petiole present in the Upper Cretaceous of the Pacific stout, 3 millimeters in length. Midrib stout, · coast on Vancouver Island and is probably prominent. Lateral primaries stout, diverging represented in the Texas remains of Oinna at acute angles from the base of the midrib, momum recorded by l{nowlton 3 from the somewhat thinner than the midrib, ascending Woodbine sand in Cooke County, Tex. It has near the n1argins two-thirds of the distance to also been recorded from the Eutaw formation the tip, where they join the basal of two or in the Chattahoochee drainage basin in Geor three pairs of short camptodrome secondaries gia. Although not known from Europe, the that are developed in that region of the leaf. forms from the Cenomanian of Bohemia which N ervilles thin, percurrent. Velenovsky describes as Aralia daphnophyl This variety is probably merely a small nar lum are very similar to this American species. row, perhaps juvenile leaf of the associated It must be obvious to any rational botanist larger-leafed species, the suppression of camp that this great range in time can scarcely rep- todrome secondaries from the outer sides of the lateral' primaries being conditioned by the nar 1 Newberry, .J. S., The flora of the Amboy clays: U. S. Geol. Survey Mon. 26, pl. 29, fig. to, 1896. rowing of the leaf. It is sufficiently distinct in 2 Watelet, A., Description des plantes fossiles du bassin de appearance to merit a name, and it rests with Paris, pl. 50, fig. 2, 1866. Sa.porta, Gaston de, Flore fosslle de Sezanne, p·l. 8, fig. 5,, 1865; Vegetation a. l'epoque des the future to verify or disprove this assump marnes heers,iennes de Gelinden, pl. 6, figs·. 5, 6, 1873 ; Re tion. vision de la fiore heersienne de Gelinden, pl. 9, figs. 2-6. 1878. Friedrich, P. A., Beitrage zur Kenntniss der Tertiar Occurrence: MeNairy sand member of Rip flora der Provinz Sachsen, pL 1, fig. 5, 1863. ley formation, Cooper pit, near Hollow Rock, 3 Hill, R. T., U. S. Geol. Survey Twenty-first Ann. Rept., pt. 7, p. 317, 1901. . Carroll County, Tenn. THYMELEALES 77 Cinnamomum newberryi minimum Berry, n. var. i meters; maximum width frmn 4 to 6.5 centi Plate XVI, figure 6 meters. Petiole stout, 1.25 to 2 centimeters in length. Midrib stout, prominent on under Leaves small, ovate, widest in the basal half side of leaf. .Lateral primaries suprabasil'ar, of the leaf and tapering upward to the acute diverging from the midrib at angles of about tip. Base broadly rounded. Length about 3.5 25° to 30°, stout and prominent below but be centimeters; maximum width about 1..2 centi coming thin distad, parallel with the leaf mar meters. Petiole stout, 6.5 millimeters in length. gins. There are two or three pairs of campto Primaries three, from top of the petiole. In drome secondaries in the distal half of the le'af general form and venation an exact miniature· and two or three sweepingly curved campto of Oinnamomum newberryi, except that the drome secondaries from the outer sides of the lateral primary on one side is craspedodrome, primaries. Tertiaries very thin, typical of the terminating in a well-marked and relatively genus, running transversely from midrib to prominent serrate tooth. Otherwise the ~ar lateral primaries and usually inosculating mid gi ns are en tire. way between the two. Texture coriaceous. This variety is probably only a reduced and This species in its size, general outline, perhaps 'abnormal leaf of the contemporaneous coriaceous texture, and coarse suprabasilar forn1 identified as Oinna11Wmum newberryi. primaries is exceedingly close to the European The fact that only a single specimen is known Tertiary spedes Oinnamomum spectabile makes this supposi~ion still more probable. Heer,4 a form that has been recorded from a Occurrence: J\1cN airy sand member of Rip large number of European localities ranging ley formation, Cooper pit, near Hollow Rock, in age from Oligocene to Pliocene and re Carroll County, Tenn. corded by Knowlton 5 from the early basic breccias of Fort Union age in Yellowstone Cinnamomum newberryi ellitpticum. Berry, n. var. Park. Another similar form, somewhat smaller Plate XVI, figure 7 in size and more delicate in texture, is Oin A relatively shorter and more robust form namomum buchii Heer,6 a widespread Eo of leaf than is typical in Oinnamomum new cene to Pliocene form in Europe and recorded berryi but not greatly different from many from the Wilcox Eocene of Louisiana. 7 It is '.fonns th'at have been referred to that species. difficult to appreciate the differences that have The fact that in the specimen here figured led European students to separate the two there were two apparently opi,lOsite .leaves at species just mentioned, for they appear to in the apex of an abortive twig may indicate that tergrade. Moreover, it is an unsettled ques the leaves are merely somewhat abnormal tion to what extent the normally lanceolate. or leaves of the type. Length about 7.5 centi ovate-lanceolate Cinnamomums assume this meters; maximum width about 3.25 centi a piculate form in their broader leaves. Cer meters. Petiole stout, curved, 1.3 centimeters tainly among the leaves of the existing Oinna in length. Primaries stout, three, slightly 'lnom1£m camphora N ees and Ebermaier the suprabasilar. Secondary venation typical ·of broader leaves tend to be apiculate and ap Oinn.anwmum. proach the fossil species enumerated above, al Occurrence: MeN airy sand member of Rip though the narro,ver leaves are ovate-lanceo ley formation, Cooper pit, near Hollow Rock, late and quite different in general appearance .. Carroll. Conn ty, Tenn. Amid the abundant materials referred to Oinnamomum ·from this and earlier Upper Cinnamomum praespectabHe Berry, n. sp. Cretaceous horizons none are at all closely Plate XVII, figures 4, 5 similar to the present species. The relation of Medium to large coarse leaves, obovate, 'Heer, Oswald, Flora tertiaria Helvetiae, vol. 2, p. 91, pl. widest above the middle, narrowing gradually 96, figs. 1-8, 1856. 6 Knowlton, F. H., U. S. Geol. Survey Mon. 32, p-. 72.7, pl. proximad to the sharply cuneate base · and 94, fig. 6, 1899. narrowing abruptly distad to the extended a Op. ci.t., p. 90, pl. 9·.'5, figs. 1-8. 7 Berry, E. W .. U. S. · Geol. Survey Prof. Paper 91, p. 299, apiculate tip. Length from 8. to 13 centi- pl. 79, fig. 10,. 1916. 78 THE FLORA OF T'HE RIPLEY FORMATION this species to the Eocene forms from American Genus MALAPOENNA Nees· rocks that have been referred to either Oin Malapoenna horrellensis Berry namomum b1.whii or Oinnamomum ·spectabile Malapoe·nna ho1·1·ellensis Berry, Torrey Bot. Club is close. The name chosen is intended to sug Bull., vol. 37, p. 198, pl. 24, figs. 1-9, 1910 ; idem, gest that· it stood in an ancestral relationship, vol. 43, p. 299, 1916. and I am inclined to think that the two Ameri can Eocene forms above mentioned are not only I described this species in 1910 as follows: identical but distinct from the essentially later Leaves ovate-lanceolate, about 8 ·centimeters long European types of tl~ose species. A recent by 2.5 centimeters in greatest. width; broadest at the evenly rounded or slightly acute base, narrowing species that is similar to all of these is Oin- gradually upward, the apex narrow and extended . namom1lm glanduliferum Meissner. but obtusely pointed. Leaf substance thin but per OccuiTence: MeN airy sand member of Rip sistent, evidently coriaceous in life, since these leaves ley formation, Cooper pit, near Hollow Rock. occur abundantly at a locality where all the vege Carroll County, Tenn. · t~ble remains except the resistant ATa1waria, Oun nmghamites, and Pistia were evidently thoroughly macerated before entombment. Secondaries four to Genus NECTANDRA Roland si.x pairs, subopposite, curved upward, camptodrome, N ectandra prolifica Berry, n. sp. branching from the midrib at an acute angle, the lowest ·pair branching from the top of the petiole Plate XVII, figures l-3 and extending upward halfway to the apex or farther giving the leaf a triple-veined appearance. Perhap~ Lanceolate, mostly falcate leaves of variable they should' be· termed lateral primaries, although size and coriaceous texture, widest below the · they are much finer than the medium-stout midrib. middle, tapering upward to an · extended The next pair of secondaries branch at a less acute angle a considerable distance above the base, one acuminate tip and downward to a narrowly third to one-half the distance to the apex. Tertiary cuneate or slightly decurrent base. Margins venation typicaHy lauraceous. entire, often somewhat undulate. Length from 8 to 12 centimeters; maximum width from 7 This species is markedly distinct from the t~ 14 millimeters. Petiole stout, about 1 centi- species of lauraceous le;:tves hitherto· described meter in length. 1\iidrib stout, prominent, in its rounded base, the only genus of this usually curved. Secondaries relatively stout, family with such a character being Oinnamo well marked, irregularly and rather remotely mum. The present species is possibly liable spaced, diverging from the midrib at angles of to be confused with Oinnamo1num heeri Les about 45° and in general ascending in sweep- quereux when. only the basal part of the leaf ing camptodrome arches; some of them diverO'- is found, but the general proportions and char ing at more open angles and pursuino· e.a acters of the whole leaf are perfectly distinct.- straighter course, particularly in the u;per It was described from several localities ·in part of the leaf. Tertiaries. thin, formin La·unts nngusta, 1-Ieer, Flora fossilis arctica, vol. 7, the Cooper pit are considerably wider than the pt. 2, p. 76, pl. 20, figs. 1b, 7; pl. 43, fig. 1c, 1882; average but are otherwise· identical. vol. 7, p. 30, pl. 57, fig. 1b, 1883 [not Rafi The species is not found in the Magothy nesque]. I,esquereux:, Flora of the Dakota group : U. S. formation of New Jersey or ~1aryland nor in Geol. Survey Mon. 17, p. 93, pl. 16, fig. 7, 1892. the· Tuscaloosa formation of Alabmna, and as Hollick, New York Bot. Garden Bull., vol. 3, p. the occurrences cited from Lesquereux and 408, pl. 70, figs. 10, 11, 1904 ; U. S. Geol. Survey Hollick are both based upon questionable evi Mon. 50, p. 81, pl. 27, figs. ·u, 12, 1907. dence it may be regarded as a form more Leaves entire, linear,..lanceolate, tapering diagnostic of the horizon of the Patoot beds · about equally to the ·acuminate apex and base. of Greenland and the Ripley of the Atlantic Length 12 centimeters or less, some of Heer's Coastal Plain. Greenland material consi.derably under this di Occurrence : MeN airy sand member of mension; width about 1.5 centimeters, although Ripley formation, Perry place, Henry County, some specimens reach 2 centimeters. Midrib and Cooper pit, near Hollow Rock, Carroll relatively thin. Secondaries thin, somewhat County, Tenn. widely separated, branching from the midrib at ·acute angles ( 45° or le~s), nine to twelve al Laurus asiminoides Berry, n. sp~ ternate pairs, finally ascending along the mar Plate XVIII, figures 1, 2• gin, camptodrome. Tertiaries forminO'b the characteristic areolat_ion of the genus. A.simina eocenicaf Kn<;>wlton, U. S. Geol. Survey Bull. 163, p. 57, pl. 14, fig. 3, 1900 [not Lesquereux, I-Ieer originally compared this species with 1878]. . ~ Laur~ts pl~ttonia, pointing out that it was more linear and acuminate. ·He also remarks Leaves broadly lanceolate, relatively small· with a similarly acuminate apex and base: that it is somewhat smaller, but this O'eneral~ • • b 1zation has been found not to hold O'Ood even reg~llarly rounded entire margins, and sub for the Greenland material, and is b especially' conaceous texture. Length 7.5 to 8 centi not true of the forms from the Middendorf meters; maxiinum width, n1idway betw·een th2 arkose member of the Black Creek formation apex and the base, 2:1 to 2.5 centimeters. jn South Carolina. and those from the Ripley Petiole stout, curved, 1 centimeter in len oth. of Tennessee. The primaries are sparser and Midrib stout. Secondaries stout, six or s:Ven more ascending than in Laurus plutonia, and J>airs, diverging from the midrib at acute an gles and sweeping upward in lonO', slio·htly ir- the more linear form gives the leaf a decidedly o b • different aspect. There can be no doubt that ·egular curves, eventually camptodrome. Ter- the two are. perfectly distinct species, natu tiaries thhi', forming open n1eshes. . rally possessing certain lauraceous characters The earliest known specimen of this species in common. Laurus .plutonia did not persist comprised . a single leaf from the Montana in the Upper Cretaceous as late as Laurus group of vVyoming which l{nowlton identified atanensis. with some doubt as Asi?nina eocenica Lesque reux.10 In describing the supposed Midway La~trus atanensis was described .from speci mens found in the Atane beds of Greenland Eocene plants from Texas in 1916 I pointe,i 11 and was subsequently recorded from · the out that the ~1ontana group leaf ·was dis·· Patoot beds. Lesquereux identified it in ma tinct from Lesquereux's type of material from terial from the Dakota sandstone of l{ansas. the Denver formation and Black Buttes, vVyo., although this record was based upon extremely both of which are now considered by l{nowl doubtful material, and Hollick has recorded ton of Eocene age. Leaves sinl.ilar to the one it from the clays of Northport, Lono- Island describ~d by l{nowIt on are sparingly repre . 0 . ' wh ICh are probably of Magothy age. Char- sented In the Ripley formation of Tennessee. acteristic material was described by me from and substantiate this conclusion. ]'hey have the Middendorf arkose member of the Black all the characters of the Lauraceae and are Creek formation in South Carolina, and iden·· uniformly smaller · than those of Asimina tical remains occur in considerable abundance 10 !.esqucn:ux. Leo, Tertiary· flora, p. 251, p. 43. figs ·.3-8 in the Ripley formation in Henry and Carroll 1878. . ' . ' 11 Berry, E. W., U. S. Geol. Survey Prof. Paper 91 p 14 counties, Tenn. Some of the specimens from 1916. . ' . • 80 THE FLORA. OF. THE RIPLEY FORMATION eocenica Lesquereux, with fewer and much camptodrome, arching near the margin and each more ascending secondaries, shorter petioles, joining the next above by a series of loops; interme and less stout midribs. diate secondaries frequent; usually joining the sec ondary next below; secondaries strong, oblique to the Occurrence: MeN airy sand member of Rip secondaries ; finer nervation not preserved. ley formation, Perry place, Henry County, The leaf here figured as the type of this spec~es is Tenn. narrowly lanceolate in shape, being. about 11 centi meters in length and 2.5 centimeters in width ; the Laurus ripleyensis Berry, n. sp. petiole is preserved for a length of 0.5 centimeter but Plate XVIII, figure 3 probably is not complete. In the matter of nervation the present species ap Leaves relatively small, ovate, widest in the proaches very closely to Laurus prilmigenia Unger, as middle and narrowing distad to a bluntly figured by Lesquereux,IZ but it differs from this in size and to some extent in shape, being much larger pointed tip, remaining full proximad and nar and narrower, especially at the base. Its thick rowing rapidly to an acuminate base. Mar·· petiole and thick basal portion of the midrib con . gins entire. Texture subcoriaceous. · Petiole stitute other minor differences. fairly long and very stout, 1.25 centimeters in It is not common in the Ripley but is a ch1ar length. Midrib stout and usually curved, acteristic form in the Vermejo formation in prominent on the under side of the leaf. Sec Colorado and the Fruitland formation in ondaries g(;out, camptodrome, about seven northwestern New Mexico and is probably pairs, those in the basal half of the leaf as present at Point of Rocks, Wyo., as l{nowlton cending subparallel with the lower lateral has pointed out. margins. Upper four pairs diverging from Occurrence: McNairy sand member of Rip· . the midrib at conspicuously wider angles, ley formation, Perry place, Henry CountJ, rather straight in their courses, conspicuously Tenn. camptodrome by a single large loop some dis tance within the margins. Tertiaries typically Genus· LA UROPHYLL UM Goeppert lauraceous, prominently seen only on impres Laurophyllum ripleyensis Berry, n. sp. sions showing the lower surface of the leaf. Length about 8 centimeters; maximum width Plate XVIII, figures 4-8 about 2.2 centimeters. Laurus coloradensis Knowlton, U. S. Geol. Survey This well-n1arked Rpecies, which is sugges Prof. Paper 98, pl. 88, fig. 5, 1916 [not fig. 4]. tive of our existing species of Persea of the Elongate-lanceolate and generally falcate southeastern United States, was apparently leaves of variable size and proportions. Tip not common in the Ripley-at lease-it is a rare generally much extended and acuminate. Base form in the collections. It is readily differenti gradually narrowed, acuminate-cuneate, or de ated from the numerous Upper Cretaceous current. Margins entire. Texture coriaceous. species of Laurtts by its general ovate form Petiole very stout, curved, thickening proxi and blunt tip. mad, 7. centimeters long in the smallest leaf, 1.5 Occurrence: McNairy sand member .of Rip centimeter~ in the la.rges~. Midrib very stout, ley formation, Perry place, Henry County, usually curved, prominent on the lower sur Tenn. · face of the leaf. Secondaries numerous, thin, Laurus coloradensis Knowlton diverging at wide angles and camptodrome, ob Lau1·us coloradensis Knowlton, U. S. Geol. Survey solete by immersion in the thick leaf substance Prof. Paper 98, p. 340, pl. 88, fig. 4 [not fig. 5]. in most specimens. Tertiaries obsolete. The 1916 ; Prof. Paper 101, p. 268, pl. 45, fig. 3, smallest perfect leaf in the collection is 4.5 1917. centimeters long by 1.25 centimeters in maxi This species was described as follows by mum width; the largest is 12 by 3 centimeters, l{nowlton in 191'7: the region of maximum width being about mid Leaf na;rowly Ianceolate, tapering in about equal way between the apex and base. A great many degree to both base and apex·; petiole short ( ?) , leaves are somewhat more slender than these, thick; midrib relatively very thick below but becom ing very thin above; secondaries numerous, a dozen 12 Lesquereux,. Leo, The 'l'ertiary flora: U. S. Geol. Survey or more pairs, subopposite, much curved upward, '!'err. Rept., vol. 7, pl. 36, fig. 6, 18i8. MYRTALES 81 and an average would perhaps be 10 centi which forms a marginal hem less than 0.5 meters in length and 2 centimeters in maximum millimeter from the margin. Texture coria width. ceous. The present characteristic and exceedingly The present species is very close to some of abundant species belongs to a type of leaf the numerous forms which from time to time which is abundant throughout most .of the Up have been referred to Eucalypt'i.t8 geinitzi per Cretaceous deposits of the world and is ( Heer) Heer. It is, however, distinct from commonly referred to the genus Laurophyllum, that species, especially when compared with which will serve the present purpose. The de Heer's type or with the more typical tails of venation that would settle its reference American material. In general it is a smaller to the Lauraceae can not be made out; conse .leaf, has a larger ·and longer petiole and an quently its true generic relations ~n the family outline less inclined toward ovate, and is rela can not be determined, although it shows a tively much more produced apically. It .is great similarity to the lower Eocene leaves typically /./yrcia-like in all its characters and from this same region which I have referred to is readily distinguishable from the forms from the genus N ectandra. the Tuscaloosa formation which have been re One of the figured specimens fr01n the Fruit ferred to Eucalyptus geinitzi. It is confined land formation of San Juan County, N.Mex.~ to the basal beds of the Eutaw formation in described by J(nowlton appears to be identical IIale County, Ala., and the Ripley formation with this species and to differ from the other in western Tennessee, and takes its name from specimens from that formation and from the the town of Havana Ala., near which the leaf type of Lau1·u,s coloradensis, which came from bearing laminated clays of the Eutaw forma the Vermejo formation.13 tion crop out. A single specimen collected by Occurrence: McNairy sand member of Rip R. T. Hill in 1891 at the big cut on the South ley formation, Perry place, Henry County ern Railway east of Pocahontas, Tenn., and (very common) ; Cooper pit, near Hollow identified by Ward as Eucalyptus attenuata Rock, Carroll County (rare), Tenn. Newberry proves to belong to this species. Occurrence: Eutaw :formation (basal beds), Order MYRTALES 2 miles south of Havana, Hale County, Ala. Family MYRTACEAE (MeN airy sand member), o:f Ripley :formation Genus MYRCIA De Candolle Camden-Paris road, 13 miles northwest of Camden, Henry County; big cut ll/2 miles Myrcia ha vanensis Berry west of Cypress, 21j2 m!les southwest of Plate XIX, figure 9 Selmer, MeN airy County ; Cooper pit, near Myrcia hava.nensis Berry, Torrey Bot. Club Bull., vol. Hollow Rock, Carroll County, Tenn. 43, p. 300, 1916; U. S. Geol. Survey Prof. Paper 112, p. 125, pl. 11, fig. 4; pl. 28, fig. 7, 1919. Myrcia dubia Berry, n. sp. E1wal1Jptus attenuata Ward [not Newberry], U. S. Geol. Surv~y Fifteenth Ann. Rept., p. 371, 1895. Plate XIX, figure 3 Leaves linear-lanceolate, falcate, about 9 cen A tiny elongate-lanceolate leaf, widest below timeters in length by 1 centime,ter in maximum the middle and tapering upward to the acumi width, which is in the lower half of the leaf. nate tip and decurrent to the petiole. Margins J\1argins entire. Apex gradually narrowed, entire. Texture coriaceous. Length about 3 acuminate. Base narrowly pointed, decurrent. centimeters; maximum width about. 4 milli Petiole very stout, tapering upward, 1.75 centi meters. Petiole expanded, about 7 millimeters meters in length. Midrib stout, curved. Sec in length. Midrib relatively very stout, ex ondaries numerous, thin, somewhat irregularly panding proximad. Stout aerodrome veins spaced, 2 to 6 millimeters apart, branching_ follow each margin, and ascending secondaries from the midrib at angles of about 40°, run connect these with the midrib. ning with but slight curvature to the well This form may represent a juvenile leaf, as marked and nearly straight longitudinal vein, it is represented in the collections by only t"Vo specimens, but its extremely coriaceous texture 1s Knowlton, F. H., U. S. Geol. Survey Prof. Paper 101, p. 268, pl. 45, fig. 3, 1917. rather suggests that it represents a small- 82 THE FLORA OF THE RIPLEY FORMATION leafed species of 111 yrcia, as the venation is Leaves linear, with acuminate tip, narrowly characteristic of that genus. It resembles cuneate base, prominent midrib, fine and close somewhat the lower Eocene . (Wilcox group) set thin ascending secondaries, .and marginal species My1•cia grenadensis Berry.14 vein. Texture, coriaceous. Occurrence: J\fcN airy sand member of ·This so-called species was instituted by Ripley formation, Cooper pit, near Hollow v elenovsky and based upon material from the Rock, Carroll County, Tenn. Perucer beds of Bohemia, and despite the fact that this author describes supposed fruits in Genus EUGENIA Linne connection with it, I have transferred it from Eugenia? anceps Berry Eucalyptus to M yrtophyllum, for there are conclusive general reasons, which I have set Plate XIX, figure 6 forth in another connection,I5 for doubting its· FJugeniaf anoeps Berry, Torrey Bot. Club Bull., vol. reference to Eucalyptus. 43, p. 301, pl. 16, figs. 2-4, 1916; U. S. Geol. There is also grave doubt as to whether the Survey Prof. Paper 112, p. 125, pl. 32, figs. references cited above all refer to the same 3-5, 1919. species. The Bohemian material shows a wide Coriaceous leaves of variable size and form, range in size and is made to include leaves lanceolate or oblong-lanceolate. Apex and which in this country are commonly referred base equally acuminate, or the apex somewhat to Eucalyptus geninitzi Heer, a second wide more attenuated. Margins entire. Length ranging and probably composite species. 7.75 to 10 centimeters; ·maximum width, mid Whatever may be the opinion as to the spe way between the apex and the base, 11 to 18 cific identity between the European and Ameri millimeters. Petiole enlarged, short and stout, can forms, there is not the slightest doubt that 3 to 4 millimeters in length. Midrib stout. the forms from the Black Creek, Eutaw, and Secondaries thin, immersed in the leaf sub ·Ripley formations are identical and represent stance. a single species of J\fyrtaceae. These occur in This species is referred to Ettgenia with con the Black Creek formation of South Carolina siderable hesitation. The material is abun and the Eutaw and Ripley formations of dant but poorly preserved, and these leaves Georgia and are exceedingly common in the resemble a variety o"f forms referred to such Ripley formation of western Tennessee. The genera as Salix and Laurophyllum. characteristic venation of this species, which E_ugenia has furnished a species in the Tus may be compared with Eocene and existing caloosa formation of Alabama and another in American species of M yrcia, is not shown in the Dakota sandstone of the weStern United the specimen figured but is well illustrated in States, and it is not uncommon in the Eocf-ne Professional Paper 84 (Pl. XIV, fig. 2; Pl. II, of the Mississippi embayment region. figs. 2-4). Occurrence : MeN airy sand . member of Occurrence: McNairy sand member of Rip Ripley formation, 2lj2 miles southwest of ley formation, Pe.rry place, Henry County ; Selmer, MeN airy County; Cooper pit, near Cooper pit, near Hollow Rock, and a quarter I-Iollow Rock, Carroll County, Terin. of a mile west of Buena Vista, Carroll County; · 2lj2 miles south of Mi:fflin, Chester County, Genus MYRTOPHYLLUM Heer Tenn. Cusseta sand 1nember of Ripley for Myrtophyllum angustum (Velenovsky) Berr)r, mation, Buena Vista, Ga. · Eutaw formation, n. comb. McBrides Ford, Columbus, Ga. Plate XIX, figure 1 Order UMBELLALES Eucalyptus ang·usta Velenovsky, Die Flora der bohm Family ARALIACEAE ischen Kreideformation, vol. 4, p. 3, pl. 3, figs. _Genus ARALIA Linne 2-12, 1885. Berry, Torrey Bot. Club Bull., vol. 36, .P· 260, pl. Aralia wellingtoniana minor Berry, n. var. 18, fig. 5, 1909; idem, vol. 37, p. 504, 1910; U. S. Plate XIX, figures 4, 5; Plate XXIII, figures 4·, 5 Geol. Survey Prof. Paper 84, pp. 55, 119, pl. 14, fig. 2; pl. 20, figs. 2-4, 1914; New J"ersey Geol. Leaves relatively small, palmately three Survey Bull. 3, p. 193, pl. 28, figs. 1-4, 1911. lobed. Lobes acutely pointed, elongate, coni- 1' Berry, E. W., U. S. Geol. Survey Prof. Paper 91, p., 316, 1s Berry, E. W., Bot. Gaz., vol. 59, pp. 484-490, 1915 ; Sci pl. 91, fig. 3, 19·16. ence, new ser., vol. 49. pp·. 91-92, 1919. UMBELLALES. 83 cal or sometimes slightly inflated, divided by plies affiliation with the earlier records of that open rounded sinuses more than three-fourths species in the same general region. of the distance to the base. Base rounded or Occurrence: McNairy sand member of Rip broadly cuneate. Margins remotely serrate. ley formation, Perry place, Henry County, Texture subcoriaceous. Length about 9 centi 'Tenn. meters; maximum width, from tip to tip of Aralia problematica Berry, n. sp. lateral lobes, about 11 centimeters. Petiole Plate XX, figures 1, 2 short and stout. Midrib stout, slightly flexu ous, prominent on the lower surface of the Fragmentary remains present in consider leaf. Lateral primaries suprabasilar, stout, able abundance and indicative of a palmately prominent, diverging from midrib at angles .of lobed or divided leaf of considerable size. The about 35 ° and curving outward. Secondanes fragments, which range in length from 10 to thin numerous, relatively straight, running to 12 centimeters and in width from 2 to 4 centi a marginal' hem, especially pronounced In. t 11e meters, are inequilateral and have the margir.s reo·ion of the sinuses. Short branches· fron1 irregularly and remotely sublobate. The tex th: marginals enter the teeth. Tertiaries thin, ture is coriaceous, and the midvein is ex mostly transverse to secondaries, forming a tremely stout, prominent,. and longitudinally roughly quadrangular areolation. wrinkled. The secondary veins are stout and prominent, they diverge from the midvein at This va.riety suggests the common Upper wide angles and irrugular intervals, and most Cretaceous Aralia 1vellingtoniana Lesquereux of them run ·straight· to a conspicuous mar in appearance but is clearly distinct. is !t ginal vein that follows all the irregularities based upon three specimens from the R1pl~y of the margin. The tertiaries are prominent, formation which indicate a leaf not over half straight, and largely percurrent. The are1· the size of the normal A. wellingtoniana. It lation is coarse and prevailingly quadrangular. lacks the decurrent base, a n1arked feature of .This species, obviously new, is unfortu·nately that species, has more spreading and les~ ~n represented by fragments, so that its characters flated and more conical lobes, more aqu1hne can not be fully determined. The mater~al teeth, straighter and less ascending second suggests a palmately lobate leaf, such as Aralia aries more open sinuses and the well-developed iJnultifida Saporta 17 or Aralia dissecta Lesque hem, especially in the region of mar~inal ~i-: reux.18 but it may have been pedately. divided, which is not developed in A. 'welling nuse~, as one fragment that appears to be identical toniana.. with the others has a petiole several millimeters A 1·alia wellingtoniana is· a characteristic in length. Both the irregular form and venation species of the Dakota sandstone and the Wood suggest certain existing leaves of the genus J a bine and Raritan formations and is hence an tropha of the family Euphorbiaceae, and it is older form than this variety. There has been of interest that Aralia multifida referred to some confusion in the past regarding its above was described in the first instance as J a limits, as Lesquereux grouped under it certain tropha pri1naeva by Saporta. On the other five-lobed leaves belonging to his Aralia sapor hand, the fragments may sjmply be those of an tana.10 irregular simple leaf, in which case I would There is some resemblance between the Rip compare them with the genus Dalechampia ley form and Aralia cottondalensis, of the Linne, a genus of the Euphorbiaceae not re Tuscaloosa formation of Alabama. The latter corded in the fossil state but with about 60 has much shorter and more conical lobesj existing species in the Tropics of both hemis broadly rounded base, and crenate marginal ph"eres. teeth. This variety also shows a similQ,rity to I have referred the present species to Aralia A 1·alia decurrens V elenovsky, a Cenomanian more as a matter of convenience rather than species from Bohemia. from a conviction that it represents a Ripley The description of the present fonn as n. 17 Sapo.t1:a, Gaston d.e, Etudes sur Ia vegetation du sud-est variety of Aralia 1oellingtoniana naturally im- de la France A. l'~poque tertiaire, vol. 1, p. 115, pl. 12, fig. 1, 1863. 10 See Berry,. E. w., To.rrey Bot. Club Bull., vol~ 39, p. 402, 18 Lesquereux. Leo, Cretaceous and Tertiary floras, p. 176, 1912. 1 pi. 35, 1863. 84 THE FLORA OF T'HE RIPLEY FORMATION form of Araliaceae, as the material is too in rib stout, prominent. Secondaries thin, about complete to serve for a new genus and Aralia ten subopposite to alternate camptodrome as used by paleobotanists is sufficiently broad to pairs; those in the lower part of the leaf di include it. My own preference would be to verge from the midrib at actute angles and refer it to the family Euphorbiaceae, but this ascend in sweeping curves parallel to lower conclusion as well as a more corp.plete descrip lateral leaf n1argins; the secondaries gradually tion must await the discovery of better ma become shorter and less ascending in the mid terial. die and upper part of the leaf. Tertiaries Occurrence: MeN airy sand member of Rip thin but well marked; below they are long~ ley formation, Perry place, Henry County, and subparallel to secondaries, with which they Cooper pit, near Hollow Rock, Carroll County, inosculate at acute angles; above they form Tenn. polygonal meshes, acute in the axis of direc Family CORNACEAE1 tion of the secondaries, forming the char acteristic areolation shoWll in the figure. Im- ' Genus CORNOPHYLLUM Newberry pressions of the lower surface indicate that the Cornophyllum minimum Berry, n. sp. underside of the leaf was clothed with stout hairs. Plate XIX, figure 2 This species is clearly allied to the group Leaves small, elliptical, widest in the middle, of somewhat earlier Upper Cretaceous form~ bluntly pointed 'and slightly decurrent to the tha~ have been referred to the genus An long petiole. Margins entire, evenly rounded, dromeda and are so common in the Atlantic and symmetrical. Leaf substance thin, char Coastal Plain, and of which Andromeda taceous. Length about 2.5 centimeters; max parlatorii Heer is probably the most common. imum width about 1.25 centimeters. Petiole Their generic reference is by no means beyond long and stout, curved, about 11 millimeters in question. The present species differs from the length. Midrib stout. Secondaries thin ; earlier forms in the general outline, the m:.t about five pairs diverge from the midrib at jority being more acuminate, and also in the acute angles, ascend subparallel with the lower thinness of the. secondaries. In outline it re l'ateral margins, and are eventually campto sembles some of the abundant lanceolate forms drome. Tertiaries where seen are percurrent. of Ficus of the Coastal Plain Cretaceous, but This is a not abundant form of uncer the venation is markedly different. tain relationship, referred to Oornophyllum Occurrence : MeN airy sand member of from its resemblance in both form and ascend Ripley formation, Perry place, Henry County, ing secondaries to the larger leaves commonly Tenn. referred to that genus. Occurrence : MeN airy sand member of Rip Order EBENALES ley formation, Cooper pit, near Hollow Rock, Family SAPOTACEAE Carroll County, Tenn. Genus CHRYSOPHYLLUM Linne Chrysophyllum parvum Berry, n. sp. Series GAMOPETALAE Plate XIX, figures 7, 8 Order ERICALES 1 Leaves small, obovate, with a cuneate base Family ERICACEAE 1 and a slightly emarginate tip. Extremely Genus ANDROMEDA Linne coriaceous, with entire, slightly revolute mar gins. Length about 3 centimeters; maximum Andromeda anceps Berry, n. sp. vdclth about 1.5 centimeters. Petiole wanting. Plate XX, figure 3 l\iiclrib very stout and prominent, curved. Lanceolate leaves of rather large size, with Secondaries numerous, stout, subparallel; nine acute tip and base, widest below the middle and or ten subopposite pairs diverge from the mid slightly more tapering distad than proximad. rib at wide angles (about 75° to 80°)' pursue Margins entire, evenly rounded. Texture sub a straight outward course, and have their ends coriaceous. Length about 18.5 centimeters; united by arches in the marginal region. Ter maximum width about 4.25 centimeters. tiaries well marked, comprising in the median Petiole missing, presumably very stout.. Mid- region veins parallel with the secondaries, EBENALES 85 which break up to. form open polygonal meshes species may also be compared with Oolutea in the ma.rgina.I region. obovata Berry,2i of the earlier Upper Creta This well-marked sma1l for1n shows points of ceous Tuscaloosa and Magothy formations, similarity with some modern species of 'Jf imu and with Oolutea protogaea Heer,22 of the sops, a genus related to Ohrysophyllum, and Patoot beds of western Greenl'and. The also with certain species of Ficus, as for ex species is represented in the present collection ample Ficus nitida Thunberg. It "might ,be by two ·specimens. considered an aberrant Fimts leaf of the Ficus Occurrence: MeN airy sand member of Rip .benja1ninea type but appears to resemble most ley formation, Perry place, Henry County, closely various species of Ohrysophyllum. It Tenn. shows some resemblance to the genus Oalyp Bumelia ripleyilna Berry, n. sp. tranthes Swartz, of the family Myrtaceae, but the venation is more nearly ,that characteristic Plate XX, figure 6 of Okrysophyllum. Leav~s orbicular, divided one-third of the Occ\nTence: J\1cNairy sand member of Rip distance to the base by an open, rather ley formation, Cooper pit, near Hollow Rock, straight-sided, prox.imally rounded sinus. Carroll County, Tenn. Ears widely separated, inequilateral, roundecl at their tips. Base broadly rounded to the Genus BUMELIA Swartz point of attachment, where the margins are Bumelia prewilcoxiana Berry. n. sp. slightly decurrent. Margins· entire, somewhat Plate XX, figure 5 ; Plate XXI, figure 4 undulate. Texture coriaceous, the surface of Leaves small, ol;>cord'ate, with a wide retuse the leaf polished. Midrib extremely stout and apex and a broadly rounded and ultimately de prominent. Secondaries thin, immersed in the current base. Margins entire, full but slightly substance of the leaf; four or five pairs di irregular. Texture subcoriaceous. Length verge from the midrib at angles of about 45° 2 to 5 centimetel's or slightly less; maximum and are camptodrome in the marginal region. width, in the upper half of the leaf, 1.75 to 3.5 Areolation a finely reticulated mesh of four centimeters. Ears broad, evenly rounded. or five sided areolae. Petiole wanting, which Sinus retuse, extending one-sixth to one-sev: suggests that the leaf was sessib~ as so thick enth of the distance to the base. Petiole re~a a midrib would scarcely be separated from a tively long and very stout, curved, about 6 petiole o:f corresponding thickness at this ex millimeters in length. Midrib very stout and act point had a petiole been present in life. prominent, curved. Secondaries stout but This is a well-marked member of the Ripley largely immersed in the leaf substance; four flora, readily distinguished from the associ or five subopposite pairs diverge from the mid ated emarginate or retuse :forms. It is a type rib at wjde angles an:d are camptodrome. Ter that has suggested the genus Oolutea to some tiaries thin, largely ·percurrent. paleobotanists but is more profitably com This well-marked species is considered to pared with Bumelia, of which there are ft have been ancestral to the lower Eocene species number o:f comparable forms in the early 19 Eocene o:f the Mississippi embayment region. Bumelia 1.vilcoxiana Berry, a~ is indicated by the name assigned to it. Bumelia wilcoxiana It may be compared with the somewhat simi 28 is a somewhat larger, slightly more elongated lar Sapotacites ettingshattseni Berry, of the. form, with more cuneate base and shorter pet Tuscaloosa formation in Alabama, which may .iole. This form also shows some resemblance possibly be looked upon as an ancestral type . to the antecedent Tuscaloosa species Sapota Occurrence: McNairy sand member o:f Rip cites shirleyensis Berry,20 which I suggested as ley :formation, Perry place, Henry County, a possible ancestor of Bumelia wilcoxiana in 'Tenn. discussing that species in 1916. The present 111 Berry,, E·. W., Torrey Bot. Club Bull., vol. 33, p. 175, pl. 8, figs. 5, 6, 1906. 10 Berry, E. W., U. S. Geol. Survey Prof. Paper 91, p. 336, 22 Heer, Oswald, Flora fossilis arctica, vol. 7, p. 43, p·l. 61, pl. 100, figs. 4, 5; pl. 107, fig. 3, 1916. fig. 1c ; pl. 62, fig. lc, 1883. 20 Berry, E. W., U. S. Geol. Survey Prof. Paper 112, p. 23 Berry, E·. W., U. S. GeoL Survey Prof. Paper 112, p. 135. 135, pl. 28, fig. 11 ; pl. 29, figs. 4-6, 1919. pl. 29, fig. 7, 1919. 86 THE FLORA OF THE RIPLEY FORMATION Order GENTIANALES Family APOCYN ACEAE Family ASCLEPIADACEAE Genus APOCYNOPHYLLUM Unger Genus ACERATES Nilsson Apocynophyllum giganteum Berry, n. sp. Acerates cretacea Berry, n. sp. Plate XX, figure 4 · Plate XXI, figure 2 Leaves small, narrowly elongate-linear, gen . Lea.ves of variable but large size, lanceolate, erally falcate, gradually narrowed to the apparently slightly inequilateral, widest in the acuminate tip and abruptly narrowed to the middle and with an acute a.pex and an equally pointed base. Margins entire. Texture sub acute base. J\tlargins entire. Texture coria coriaceous. Petiole short and stout, practi ceous. Length about 22 centimeters; maxi cally wanting. Midrib stout, curved, relatively mum width about 5 centiineters. The smallest prominent. Secondaries thin, numerous, as specimen seen is 15 centimeters in length and cending either obviously camptodrome or unit 3.75 centimeters in maximum width. Petiole ing to form a marginal vein. Length about 8 missing. Midrib stout, slightly flexuous, centimeters; maximum width, in the basal part pron1inent on the lower surface of the leaf. of the leaf, about 3 millimeters. Secondaries thin, numerous, and subparallel; This handsome slender species is much like they diverge from the midrib at. angles of Acerates veterana Heer, of the Arctic and about 65° at regular intervals of about 6 or European Tertiary. It is also much like cer 7 milli1neters, sweeping upward in regular tain European Te.rtiary forms that have been curves and arching along the ma.rgins. Terti referred to the genus Echitonium of the allied aries thin, prevailingly obliquely p~rcurrent, family Apocynaceae. The venation is that generally with an intermediate vein in the characteristic of Acerates and of certain exist central region of the leaf midway between the ing species of Asclepiadaceae and Apocynaceae. adjacent secondaries, with which it is subpar This species is not cornmon in the Ripley but allel. is an exceedingly characteristic form readily In general form these leaves resemble Mag recognizable even when fragmentary if the nolia lanceolata Lesquereux, but the venation venation can be seen. is diiferent and typical of A.pocynophyllum, The genus Acerates has seldom been recog although there is some resemblance to certain nized in fossil floras. Somewhat larger species fossil forms that have been referred to the than that of the Ripley ..are Acerates arctica genus Ficus. Heer, 2 ~ from the Atane and Patoot beds of ·The genus Apocynophyllum is a form genus 25 Greenland; Acerates amboyensis Berry, of for members of the family Apocynaceae whose the Tuscaloosa, Raritan, and Black Creek for generic position is uncertain and which -resem Inations of the Atlantic Coastal Plain; and ble Thevetia, Oerbera, Apocynum, and other Acerates veterana I-Ieer,26 of the European and existing genera. A considerable number.of fos Arctic Tertiary. Cockerell 27 has referred a· sil species have been described, including five follicle from the Florissant Miocene to this or six fron1 the Upper Cretaceous and over a genus. score from the Eocene of both Europe and Occurrence: J\icN airy sand member of Rip-· America. The genus has not been rec.ognized in ley formation, Cooper pit, near Hollow Rock, the earlier Upper Cretaceous of the Atlanti0 Carroll County, Tenn. Coa.stai Plain but is present in the Dakota u Heer, Oswald, Flora fossilis arctica, vol. 6, pt. 2, p. 82, sandstone of the vVest. It is abundantly repre pl. 30, figs. 19, 20, 1882. se~ted in the Eocene of the Mississippi embay. 211 Berry, E. W., Torrey Bot. Club Bull., vol. 36, p. 263, ment region, however. 1909. 20 Heer, Oswald, Flora tertlaria Helvetlae, vol. ·s, p. 20, pl. Occurrence: MeN airy sand member of Rip 104, figs. '5-8, 1859. ley. formation, Cooper pit, near Hollow Rock. 27 Cockerell, T. D. A., Am. Naturalist, vol. 42, p. 580, fig. 10, 1908. Carroll County~ Tenn. GENTIANALES 87 Apocynophyllum perryensis Berry, n. sp. rib, and the details of the areolation, as well Plate XXI, figure 1 as in the extremely elongate form of the leaf Oblong obtuse leaves of considerable size, as a whole. On the other hand, it agrees with widest above the middle, with a broadly round various fossil forms of Apocynophyllum and ed tip and extended tapering base. Margins with' several recent genera of the family entire, somewhat undulate. Texture ~oria A pocynacea~. ceous. Lamina inequilateral and somewhat fal The fragments from the Ripley of Alabama cate. Length about .16 centimeters; maximum (U. S. Nat. Museum catalog Nos. 34957, width about 2.5 centimeters. Midrib stout, 34958) are now seen to be referable to this prominent, and curved. Secondaries apparent species. · ly not differentiated, the venation consisting Occurrence : MeN airy sand member of solely of a· very fine isodiametric areoloation Ripley formation, Cooper pit, near Hollow suggestive of the Proteaceae. Rock, Carroll County, Tenn. Ripley forma This peculiar species iS' based upon the single tion, Cowikee Creek, Barbour County, Ala. specimen figured. lt is distinctive enough for Apocynophyllum ripleyensis Berry, n. sp. stratigraphic assignment, but the features are not sufficiently developed for comparisons with Plate XXI, :figure 3 the existing genera of Apocynacea~. Leaves linear-lanceolate, widest in the lower Occurrence:· McN~airy sand member of Rip halt, gradually narrowed upward to the great ley formation, Perry place, Henry County, ly extended and often falcate acuminate tip Tenn. and downward to the decurrent· base. Mar Apocynophyllum sumterensis (Berry) Berry gins entire. Texture subcoriaceous. Length Q1tero·u.s 81tmte1·ensis Berry, U. S. Geol. Survey Prof. from 13 to 20 centimeters; maximum width Paper 84, p. 35, pl. 10, figs. 9, 10, 1914. from 1.1 to 2. centimeters. Petiole short and " Undeterminable leaf," Berry, U. S. Geol. Survey stout. Midrib stout, prominent. Secondaries Prof.. Paper 112, pl. 31, figs. 5, 6, 1919. thin, obsolete by immersion in n1ost of the Leaves large, linear, acuminate, coriaceous, · specimens, diverging at wide ·angles at regular entire, with a broadly cuneate base. Length intervals of about 3 millimeters, and campto about 15 to 16 centimeters; greatest width 3 drome. Tertiaries thin, forming an open lat- centimeters. Petiole short and stout. Midrib . erally elongated ·polygonal areolation. stout. Secondaries thin, branching from the. This species is perhaps only a variety of midrib at angles ·of 45° ·or more, taking a Apocynophyllum s~tmterensis Berry, with rather straight course, almost to the margin which it is associated. It has the same gen and then turning sharply, upward. Tertiaries eral form but varies to a smaller size, is rela obsolete, as are also the secondaries in some tively narrower, with a different disposition of specimens. the more numerous secondaries. It is also a The Ripley material is fully as long as the more abundant form in the collections. In type material but prevailingly narrow, so that specimens lacking the venation it is liable to the ba$e is more narrowly cuneate and the ·be· confused with various linear-lanceolate secondaries are slightly more ascending. The leaves of severaJ genera such as 11/ yrcia and extremely elongate and usually somewhat fal N ectandra, but it is in general more elongated ·ca.te form is very chara~teristic. than any associated. forms. A.lthough this species was originally com Occurrence: MeN airy sand mmuber of pared unhesitatingly with the willow · and Ripley formation, Cooper pit, near Hollow hwrel oaks, this comparison 'no longer seems Rock, Carroll County, Tenn. valid. The type material was preserved in a clay ironstone and showed less detail than that Apocynophyllum alatum Berry, n. sp. preserved in the fine-grained Ripley clay. The Plate XXIII, figure 7 venation proves to differ decidedly from that of Quercus in the more close-set thin regular Elongate-lanceolate leaves with acuminate secondaries, the lack of prominence of the mid- tips and rounded decuri·ent bases expanding 88 THE FLORA OF THE RIPLEY FORMATION to winged petioles. . Margins entire, conspicu Genus· :PHYLLITES Brongniart ously undulate in smne ·specimens. Leaf sub Phyllites hydrilloides Berry n. sp. stance thin. Length from 13 to 15 centime ters; maximum width, below the middle, 1.75 Plate XXII, figure 2 to 2.25 centimeters. Petioles stolJt ana ex A small and delicate form suggestive of an panded proximad, about 1.25 c~ntimeters in aqua,tic habitat, oblanceolate, widest above the length, curved, bordered by wings increasing middle, decurrent to the sessile base. Margin in width from the sinus at the base of the with tiny remote teeth. Midrib thin, greatly lamina to the truncate(y rounded base, where expanded at the sheathing base. Rest of the they are 1 to 2 millimeters in width, making venation obsolete. Length about 1.9 centi the total width at the base of the petiole about meters; maximum width about 4 millimeters. 6 millimete.rs. Midrib stout, curved.· Second This. tiny form may be a juvenile leaf of aries nun1erous, thin, camptodrome. · Terti some of the associated lanceolate-toothed aries thin. Areolation minute. margined Ripley species, but it does not ap This form is at once distinguished by its pear to be, and its delicate texture and lack thin texture and its much expanded and con-· of venation suggest its relationship to such a spicuously alate petioles. modern form as the genus H ydrilla L. C. Occurrence: MeN airy sand 1nember of Rip Rich, of the family Hydrocharitaceae. This ley formation, Perry place, Henry County, genus includes a wide-ranging water plant of T6nn. the sea margins found in Europe, Asia, Africa, :POSITION UNCERTAIN and Australia and generally considered Genus CALYCITES of authors · monotypic, although this is very doubtful. In Calycites ripleyensis Berry, n. sp. the absence of conclusive proof of such a rela tionship the. fossil form is referred to the non Plate XXII, figure·l committal genus Phyllites. A scarious calyx of considerable consistency, Occurrence: MeN airy sand member of consisting of five triangular lanceolate sepa.ls · Ripley formation, Perry place, Henry County, coalescent proximad. The whole is about 15 Tenn. millimeters in diameter; each sepal is about 6 millimeters in length and 3 millimeters wide Phyll~tes hydrocharitoides Berry n. sp. at the base. There is no trace of any fruit Plate XXII, figure 9 · in connection with the calyx. This well-marked floral organ is represented . Elongated, slender, linear, forking foliar or by a single specimen. It suggests S olanites gans, fiat, of slight consistency but with a or Sym,plocos but ·might well represent a prominent median ~ein or midrib. No other member of almost any gamosepalous family, veins can be discerned in the lamina, which is for example, Diospyros, of the Ebenaceae. It about 2 millimeters in width. is not characteristic of Diospyros, however, These rem·ains are oom1non at the Carroll although it is identical with what Saporta County locality, and massed fragments suggest calls Diospyros rugosa, a coll!mon form in the some gymnospermous genus such as Dicrano Oligocene of southeastern France.28 Leaves phyllum, Trichopitys, or Ozekanowskia, but of Diospyros are abundant in the American more complete mater\al shows these suggestions Upper Cretaceous, although none have been to be remote from the facts. Undivided speci recognized in the Ripley, and a very distinc mens as much as 13 centimeters ih length have tive calyx occurs in beds as old as the Raritan been observed and they are too loax to be even formation. 29 considered as Cyperaceae. In two specimens Occurrence: MeN airy sand m~mber of there is an aureole of fiabellate. bra,nching a p Ripley formation, Perry place, Henry County, pendages at the node or fork, suggestive .of Tenn. flattened· rootlets from a branched fla.ttened 0 rootstock. They appear to me to be remotely 28 Saporta, Gaston de. Etudes sur Ia veg~tation du sud-est de Ia France A. l'epoque tertiaire, vol. 1, p. 111, pl. 11, fig. 3, comparable to Ph.yllospadix or T1 allisneria or 1863. some related genus, and I would suggest that 20 B~rry, E. W., Torrey Bot. Club Bull., vo·l. 38, p. 418, pl. 19, fig. 5, ] 911. they represent some water weed of the Ripley POSITION UNOERTAIN 89 sea margin referable ·to the Hydrocharitaceae Carpolithus perryensis Berry, n. sp. or Potamogetonaceae. Plate XXII, figure 3 Occurrence: MeN airy sand member of Rip ley formation, Cooper pit, near Hollow Rock, Turbinate fruits with a stout peduncle, en Carroll County, Tenn~ larged seed-bearing distal portion, and apieu Phyllites ripleyensis Berry, n. sp. late tip, the whole about 7.5 millimeters in length and 3 millimeters in maximum di-. Plate XXII, figure 4 ameter. Medium large leaves, ovate, widest in the These remains are somewhat suggestive of middle and about equally pointed •at both ends, the Upper Cretaceous genus P1·otodammara, with entire, evenly rounded margins. Texture abunda.nt at earlier horizons in the Coastal subcoriaceous. No traces of venation except Plain. They lack resin canals, however, are the midrib and a thin, fine nondistinctive areo narrower and more spherical distad, and ap lation. Length about 8.5 centimeters; maxi pear to conta.in seeds in this enlarged portion. mum width about 4.2 centimeters. It is believed that they are not coniferous but This form is without geologic or botanic sig that they represent some dicotyledonous form. nificance except that it indicates the presence Occurrence: McNairy sand member of Rip ley formation, Perry place, Henry County, of an otherwise unknown member of the Rip Tenn. ley flora. It is remarkable. in showing no traces of a secondary venation, which should Carpolithus carrollensis Berry, n. sp. be pronounced in a. leaf of th!s .size. Occurrence: McNairy sand member of Rip Plate XXII, figure 6 ley formation, Cooper pit, near Hollow Rock, A bluntly fusiform, fibrous, ligneous fruit of Carroll County, Tenn. unknown botanic affinity, 3 centimeters in Carpolithus ripleyensis Berry, n. sp. length and 1.5 centimeters in ma-ximum diam eter. Plate XXII, figure 5 These poorly characterized fruits ~re, how A relatively large drupaceous fruit, nea.rly ever, of great importance, as they appear to 4 centimeters in length and 2.8 centimeters in be identical with those from the lignites of maximum diameter as preserved in the shell Fuve~u, in southern France, described by marl of the Ripley formation at Coon Creek, Saporta 30 and con1pared by him with the nuts Tenn. Consisting of a lignified sarcotesta of of Nipadites, the Eocene representative of the great thickness and a,. hard, smooth sclerotesta existing nipa palm. The former fruits come or stone. Botanic affinity unknown, and the from a horizon usually considered of Aturian specimen is interesting chiefly from the fact age, and if the comparison of the Tennessee that it must ha.ve originally possessed con and French fruits is worth anything-and it siderable buoyancy, which is unusual in a fruit seems to me to be a legitimate and close one of this kind unless the surmise,_ that it was it is of considerable importance for .purposes drupaceous is erroneous and it was an in of correlation. It would seem that Saporta dehiscent ligneous capsula.r fruit. The fact was justified in considering the French fruits that the supposed stone is somewhat excentric. those of some palm, and I believe that the same in position renders this possible and suggests relationship is indicated for Carpolithus car that the stone is a large hard seed and that rollensis. more than one was originally present around Occurrence : MeN airy sand member of the central axis. It .should, perhaps, be com Ripley formation, Cooper pit, near Hollow pared with members of the Sapotaceae. Rock, Carroll County, Tenn. Occurrence: MeN airy sand member of Rip tey formation, Coon Creek, MeN airy County, 30 Saporta, Gaston de, Etudes sur Ia v~getation du sud-est de la France it l'epoque tertiaire, vol. 1, .P· 47, pl. 1, fig. 6, Tenn. 1863. 90 THE FLORA OF THE RIPLEY .FORMATION Halymenites major. Lesquereux Typical material is plentiful in the Coffee Halymenites major Lesquereux, Tertiary flora, p. 38. sand member of the Eutaw formation and also pl. 1, figs. 7, 8, 1878. at different horizons in the Ripley formation These objects, frequently considered to rep in western Tennessee. Generally, however, resent fossil fucoids, are very abundant in the these objects are very friable, and the collected western United States in sandy beds ranging in age from the Colorado group· to the Eocene. specimens represent only the beds enumerated They were long thought to be typical of the below. Fox Hills sandstone, in which they are abun- Occurrence: Coffee sand member of Eutaw dantly developed, but have now been found in formation, Coffee Bluff, H'ardin County, Tenn. both older and younger beds and ch'aracterize MeN airy sand member of Ripley formation, sandy ferruginous sediments. MeN airy County, Tenn. PLATES 91 INDEX A Page. Page Doryanthi tes cretacea Berry ______36-37 Acaciaphyllitcs cretaceum Berry, n. SP------57, Pl. X Dryophyllum gracile Dobey______46 Acer cretaceum Berry, n. SP------66-67, Pl. XIII protofagus Berry, n. SP------46-47, Pl. V Acerates cretacea Berry, n. SP------·----- 86, Pl. XX Dryopterites stephenson! Berry_------28 Alismaphyllum cretaceum Berry, n. sp______35-36, Pl. III Andromeda auceps Berry, n. SP------84, Pl. XX E novaecaesaroao Berry.------__ 62 Eucalyptus augusta Vclenovsky ____ ------82 A nclrovettia ______.... ______... --.----...... -- __ 32 attenuata Ward.______81 Apocynophyllum alatum Berry, n. sp ______87-88, Pl. XXIII Eug~nia'? anceps BerrY------82, Pl. XIX giganteum Berry, n. SP------86, Pl. XXI Euphorbiophyllum antiquum Saporta and Marion....•• 62,-63, Pl. XU perryensis Berry, n. SP------·------87, Pl. XXI petiolatum Berry, n. SP------62, Pl. XII ripleyonsis Berry, n. SP------87, Pl. XXI tennesseensis Berry, n. s·p ______.f- ____,.., ______62, Pl. XII sumtorensis (Berry) BerrY------87 Eutaw flora, comparison of, with the Ripley flora ______18-19 Aralia problematica Berry, n. SP------83-84, Pl. XX Eutaw formation, distribution oL------2 wollingtoniana minor Berry, n. var ------82-83, Pis. XIX, XXIII. Amucaria bladononsis Berry .. --- __ ------.-----... --.----. __ .. ___ 33-34 F jofTroyi Berry ___ ------____ ----- ______.--~- 34 Fagus ripleyonsis Berry, n. SP------47, Pis. V, VII Artocarpus crotacoa Derry, n. SP------48-50, Pis. VIII, IX Ficus bockwithii Lcsquereux______50 :Asimina ooconica '? Knowlton______79 carrollonsis Berry, n. sp ______: ______53, Pl. x· Asplenium caloptoris (Deboy and :Ettingshauson) lloor. --- 25-2C., Pl. I celtifolius DerrY------52, Pl. X D cooporensis Derry, u. SP------52, Pl. X Bnuhinin ripleyonsis Berry______58 crassipes (Ifoer) Heor.. ______51, Pl. VII Bcnizia cnloptel'is ______------______. __ . ----.. ----.. -----.. __ -. 25 georgiana Berry______53 kransiana 1Je01· ... ______,. ______50-51 Black Creek flora, comparison of, with the Ripley flora._"------19 leci Knowlton .. ______.12 .Bumelia prewilcoxiana Berry, 11. SP------85, Pis. XX, XXI obtusa-sossilia Berry, n. sp ______51, Pl. VII ripleynna Berry, 11. SP------85, Pl. XX ripleyana Berry, n. SP------50, PI. VII c suspecta Velenovsky ______------. ______.10 France, comparison of Cretaceous flora of, with tho Hipley flora.. 23 Cacsnlpinitcs penyensis Berry, n. SP------58, Pl. X Frenelites reichii Ettingshauseu______29 riployensis Derry, n. SP------58, Pl. XI Ct\llitris... _------.------.. ------.---- 29 G Cnlycitcs ripleyensis Berry, n. SP------88, Pl. XXII 9einitzia formosa Heer ______32-33 Cappnris proceuica Berry, n. SP------56-57, Pl. X gracillima Jeffrey ______------_------. 32 Carpollth\lS carrollensis Berry, n. SP------89, Pl. XXII Geonomites schimperi Lesquereux ______37-39, Pl. II perryensis Berry, n. SP------89, Pl. XXI~ rlployensis Derry, n. SP------89, Pl. XXII visianii Berry_------____ .. ______------39 Codrolt\ prowilcoxiana Berry, n. SP------61, Pl. XII Germany, comparison of Cretaceous flora of, with the Ripley flora. 22-23 Colastrophyllum cnrolinonsis Berry_------·------63-64 Gloditsiophyllum aristatum Berry, n. SP------.61, Pl. XI cnssiqoidos Derry, n. SP------65, Pl. XII preovatum Berry, n. SP------6Q-61, Pl. XI minimum Derry, n. SP------64-65, Pl. XU prooocenicum Derry, n. SP------~------(H, Pl. XI Gosau flora, comparison of, with the Ripley flora______23 porryi Berry, 11. SP------65, Pl. XII riployanum Derry, 11. SP------64, Pl. XII Grewiopsis inoquilatoralis Berry, n. ·SP------· 73, Pl. XV ripleyensis Berry, n. SP------72-73, Pl. XV variabilis Berry, 11. SP------65-66, Pl. XIII Gymnogrammo bohemica... ______26 Celtis crotacoa Berry, 11. SP------47-48, Pl. VII Chrysophyllum parvum Berry, 11. SP------84-85, Pl. XIX H Clnnamomum intormodium Smith------~-~- 75 Halymonites major Lesquoreux______90 11ewberryi DerrY------75-76, Pl. XVI Heor, Oswald, cited ______.____ 32 olliptioum Berry, n. var _------77, Pl. XVI lanceolatum Berry, n. var______76, Pl. XVI J Japan, comparison of Cretaceous flora of, with the Ripley flora.... 22 minimum Berry, n. var·------77, Pl. XVI Juglans similis Knowlton. ______•• 39 praespectabile Berry, n. sp. ------c----- 77-78, Pl. XVII sezannense Hoer •••. ------75 tennesseensis Berry, n. SP------39-40, Pl. III· Cissites crispus Velenovsky______70 wadii Berry, n. SP------~------40, Pl. IV panduratus Knowlton ••••.•••• _---- •.••• ______••••••• 7Q-72 K Coffee sand member, distribution oL...... 2 Knowlton, F. H., cited .•••••••••••••••••••••••••••••• 25, 34, 39, 46, 52,80 Colorado group, flora of. •••••••••.•.•••••••.••••••••••••••••.••·•• 21 Cornophyllum minimum Berry, n. sp..••••••.••••.•.••••• 84, Pl. XIX L Correlation of the Ripley flora •..• ---········-----·------·------23-24 Laurophyllum ripleyensls Berry, n. SP------~------80-81, Pl. XVIII Laurus angusta Hear ___ •. ______.______79 Cunninghamites elegans (Cordn) Endlicher _------33 asiminoides Berry, n. SP------79-80, Pl. XVIII D atanensis Berry ______78-79 Dalborgia cretacoa Berry, n. SP------59, Pl. XI coloradensis Knowlton .. ______80 porryana Derry, n. sp. _------59, Pl. XI riploycnsis Berry, n. SP------80, Pl. XVIII prowllcoxiana Berry, n. SP------59-60, Pl. XI Leguminosites canavaliodes Berry______60 Dammara aclcularis Knowlton ______34, Pl. XXII carrollensis Berry, n. SP------60, Pl. XI Dobey and Ettii!gshause11, cited ______25,26,27 perryensis Berry, n. sp ..... : ... ------.------60, Pl. XI Dlllonitos crotaceus Berry, n. SP------74, PI. XV Lesquoreux, Leo, cited .• __ ---- ______--- ______------____ 37 Dloscoritos crotaceus Berry, n. SP------36, Pl. III Liriodendron laramienso Ward.------53-55, Pis. XI,.XXIII 100913-25t-9 93 94 INDEX M R Page ~age Magnolia capellinii Heer. ------___ ------______55-56 ·Raphaelia minuta Berry n. sp ______·______26, Pl. .I Magothy flora, comparison of, with the Ripley flora ______19-20 neuropteroide.s Debey and E ttingshausen ______26, Pl. I Malapoenna horrellensis Berry ______:______78 Relations of Cretaceous deposits in the eastern Gulf region______1-3 Manihotites georgiana BerrY------63 Rhamnus ripleyensis Berry, n. SP------~------67, Pl. XIII Menispermites variabilis Berry __ ·------, 56 Ripley flora, comparison of, with other Upper Cretaceous floras __ 15-23 Mimosites cooperensis Berry, n. SP------~- 57, Pl. X composition, origin, and evolution oL·------~------5-7 Monheimia aquisgranensis Debey and Ettingshausen ______27, Pl. I distribution oL·------__ ------____ _ 7-15 Montana group, flora oL ______21-22 number of species, genera, etc .... ------~------12-14 Moriconia Debey and Ettingshausen·------"------. 29-30 . occurrence of..______!i americana Berry ______31-32, Pl. III Ripley formation, distribution oL ______-~ ______cyclotoxon Debey and Ettingshausen.,----~------.3D-31, Pl. III limits of. ___ .:·.·--______3 Myrcia dubia Berry, n. SP------~------81-82, Pl. XIX lithologic character oL __ ------. havanensis BerrY------~·------81, Pl. XIX stratigraphic-relations of.. ______-~- ______.,_ 3-4 Myrica ·brittoniana Berry ______c_ •. ______45, Pl. VI subdivisions of. ______------______7 brittoniana obtusata Berry, n. var ______45, Pl. V cooperensis Berry, n. SP------42-43, Pl. VI s johnstrupi (Heer) Berry ______-~---__ 43 · Sabalites sp. Berry------~---______39 Salix gardneri Knowlton? ______46, Pl. V minor Berry, n. SP------.43-44, Pl. VI Scope .of the report.- __ ------_-~- _____ -----______1 ornata Berry, n. SP------"----- 44, Pl. VI Selaginella laciniata Lesquereux •. ______.: ______2!i ripleyensis Berry ______~ ______41-42 torreyi Lesquereux. ------______~ ______.44, Pl. VI Selma chalk, distribution oL·------213 obtusata Berry, n. var ------·------.45, Pl. VI Sequoia.gracillima Newberry... c •••••• ------. 32 reichenbachi. Lange __ • ______-~______-32 wadii Berry, n. SP------40-41, Pl. IV minor Berry, n.·var ______, ______41, Pl. IV Sterculia snowii tennesseensis BerrY------73-74 Myrtop~yllum angustum (Velenovsky) Berry, n. comb .. 82, Pl. XIX T N 'l'aeniopteris SP------27-:"28, Pl. I Nectand.ra prolifica Berry, n. SP------78, Pl. XVIl Ternstroemites cretaceus Berry, n. SP------75, Pl. XVI ripleyensis Berry, ri. SP------74, Pis. XV, XVI p tennesseensis Berry, n. SP------·-- 74-75, Pl. XVI Pachystima? cretacea BerrY------66 Thinnfeldia n. sp. BerrY------··------28 Patoot flora, comparison of, with the Ripley flora ______2Q-21 Tombigbee sand member, distribution oL______2 Pecopteris kudlisetensis Heer. _____ ------30 Tuscaloosa formation, distribution oL------2 Phyllites hydrilloides Berry, n. sp ______88, Pl. XXII hydrocharitoides Berry, n. SP------88-89, Pl. XXII v ripleyensis Berry, n. SP------89, Pl. XXII Vancouver Island, flora oL------21 Platanus ripleyensis Berry_------~------53 Vermejo flora, comparison of, with the Ripley flora______21 sp. Berry____ ------______------______------___ :------__ 53 w · Portugal, comparison of Cretaceous flora of, with the Ripley flora.. i2 Potamogeton ripleyensis Berry, n. sp ______34-35, Pis. III, XXIII Widdringtonites reichii (Ettingshausen) Heer______29 Proteciides crassipes Heer·------51 z Protophyllocladus lobatus BerrY------28-29 Zizypb,us ~aurifolius BerrY------67-68, Pl..XIII Q perryi Berry, n. SP------69, Pl. XIII Quercus sumtcrensis Berry ______c_____ 87 ripleyensis Berry, n. SP------68-69, Pl. XIII 0