UE924 .B425 Berry, Edward Wilber The flora of the woodbine sand at Arthurs Bluff, Texas DEPARTMENT OF THE INTERIOR Albert B. Fall, Secretary United States Geological Survey George Otis Smith, Director Professional Paper 129 — G THE FLORA OF THE WOODBINE SAND AT ARTHURS BLUFF, TEXAS BY EDWARD WILBER BERRY Published March 23, 1922 Shorter contributions to general geology, 1921 (Pages 153-181) WASHINGTON GOVERNMENT PRINTING OFFICE 1922 f^H2J^ CONTENTS. IntitHluction 153 The Woodbine sand 154 Name 154 Character of materials 154 Fauna 155 Flora 155 Age of the flora 156 Relation to the floras of other Upper Cretaceous formations of the Coastal Plain 156 Relation to Bingen sand flora 156 Relation to Tuscahwisa flora 157 Relation to floras of other formations of the (^astal Plain 157 Relation to the flora of the Dakota sandstone 158 Relation to the flora of the Cheyenne sandstone 158 Relation of Woodliino flora to floras of other areas 158 Systematic account of the flora 159 njJTSTRATIONS. Paw. Plates XXXVI-XT/. Fossil plants fnmi tho WiHidliine sand 181 FiriURE 11. Cut in the Woodbine sand near .\rlhiirs Hluff, l.amar County, Tex 180 II THE FLORA OF THK WOODHINK SAND AT AHTHl H8 BLUFF, TEXAS. Bv EmvAitii Wii.iiKK I'.KRin. INTRODUCTION. Tho protfoiui' of fossil plants in tin,' (.'ii-ia- ceous strata of Laniar County, Tex., has boon known for over half a coiilui'v, as is shown hv a lettiT from Dr. B. F. Shuniard ilati'd Oflober 2, 1S60, read before the Academy of Science of St. Louis at its session of November o, INfiO.' Dr. Sliumard, at that time State geologist of Te.\as, states that his brother, Dr. G. G. Shumard, dis- covered in tl'.e Cretaceous yellcnvish sandstones of ] .amar Count j- near lied Ki ver numerous dico- tyledonous leaves resembling the modr-rn leaves of Salix. Ih.r. Laurus, etc. These fossils were undoubtedly from the locality now known as Arthurs Bluff (fig. 11, p. 180), which has fur- nished most of the subsecjuent collections. Dr. Shumard further states that these plants were sent to Leo Lesquereux for determination, but if sent they were apparently lost in transit. - ■ When R. T. Hill took up the study of the Texas Cretaceous, new collections were made between ISSO and 1^S5 at Arthurs BlufT and at Denison, the latter a localit}' originally ihscuv- ered by Dr. Shumard. These collections were, according to Hill, sent to the L'nited States National Musemii and lost in storage. Finally, in Hill's great work on the Texas Cretaceous," F. H. Knowlton furnished a report on collections of fossil plants from the Woodbine sand at Arthurs Bluff, I.,amar County; Woodbine, Cooke County; and Denison, Grayson County. The largest of these collections is the one from .\rthurs Bluff, which was made in 1S94 by T. Wayland Vaughan. The plants are preseixed in a fragmentary state in a yellowish ferrugi- nous sandy clay or loose sandstone. Knowl- ton identified the following species from this locality: .\ralia WfUinptoniaiia vaughanii Knowlton. Benzoin venustum (Lesquereux) Knowlton. > Acad. Sci. St. Louis Trans., vol. 2, p. HO, 1S6S. » Lcsquereirx, Leo, Cretaceous flora: V. S. CJcol. Survey Terr. Repl.. vol. 6, p. n, 1874. •Hill, R.T.,Geographyand geology ot the Black and Grand prairies, Te.x.: U. S. Gcol. Survey Twenty-first .\nn. Kept., pt. 7, pp. 314-318, pi. 39, 1901. I 'itwpyros primaeva Ileer. i''uiia f;lai*ciieana Ix-Bquereux? I.iriotlendron iiinnati(i. pp. 3'-7- 406, pis. 30-32, 1912. i.'5:i 154 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY, 1921. t^ucalyptus geinitzi ^lleer) Hecr. Ficus ilaiilinoircnoidfs (Ilpor^ Berry. Laurophyllum mimis Nowborry. Laurus plutonia lli-cr. Liriodendron qucrcifolium Xewlierry. Magnolia si)ocio8a Ilecr. Malapocnna falcifolia (I^osquorouxt Knowlton? Myrica oinarsrinata Ilcor. Orocxlaphiicala Iiamensis Horry. Palaoocassia laurinea Lcsiiucrciix. PodozamiU's lanceolatus ^Lindley and llutlon) Braun. Populus harkoriana I^esquereux. Rhamnus tenax lA's<|Uoreux. Rhus redditiformis Berry. Sapindus morrisoni Heer. Sterculia lugubris I.esquereux? Tricalycitos papyracous Xcwberry. Viburnum robustum I>^8(|ueroui? Zizyphus lamaronsis Berry. It was deemed desirable, in connection witli the problem of the age of the Dakota sandstone and the rehition between the Upper Cretaceous formations of the Coastal Plain and those of the Western Interior, that the flora of the Woodbine sand should be critically reviewetl. Consequently all the material in the United States Xational Museum from Arthurs Bluff was sent to me anil constitutes the basis of the present report. I have never visited the region, nor have I seen the collections from Cooko and Grayson counties, hence recorded forms from these locahties are ignored imless they are present in the collections studied. The total numVjer of species in the -Vrthurs Bluff mate- rial amounts to 43, and it is probable that de- tailed explorations would double or treble the number of known forms. Hence the present report must be regarded entirely as a prelim- inary contribution. Fortunately the conclusions to be derived from a study of this small flora are so complete and decisive as regards the age and relatioiisliip of the deposits that no subsequent additions to the flora can change the general conclusions, and this fact must be the excuse for calling an account of but 43 species from a single locality the flora of the Woodbine formation. THK WOODBINK SAND. NAME. The name Woodbine formation was pro- posed by Hill" in 1001 from the town of Wood- bine, in the northeastern ])art of Cooke Cotmty, • Hill, R. T., op. dt., p. SM. Tex. On account of the predoimnanceof tuicon- soliilated sand the formation is now called the Woodbine sand. These beds have a somewhat complicated nomenclatorial history. Ilill some years earlier" called them the '"Timber Creek group" or "Lower Cross Timbers formation." In still earlier years they had usually been con- sidered of Tertiary age. They were originally described by G. G. Shumard as the "arenaceous and marly clay or Red River group'' and re- ferred to the Tertiary. His brother, who was the first to record fossil plants from thi^se strata, placed them in various positions in the sections of Texas formations whicii lie pub- lisheil at different times. Hill clearh- recognized their equivalence with the Dakota sandstone but quite rightly objected to the indiscriminate use of that term and fortunately decided to apply the local name. CHARACTER OF MATERIALS. The Woodbine materials are largely cur- rent-bedded sands, generally white and friable where nonfcnuginous, in places browniish and consolidated b}"^ iron oxide, containing exten- sive to small lenses of more or less carbona- ceous laminated clays with some interbedded layers of lignite or lignitic clay. They give rise to sandy soils strewn M'ith fragments of ferruginous sandstone and siliceous ironstone. The limits of the Woodbine are uncertain. It is said to be unconformable with the under- l3'ing Denison formation of the Washita group and to pass without a break into the overlying Eagle Ford formation. TafT " described the formation under the name Dakota sand and divided it into Timber Creek beds. Dexter sand, anr sand mem- ber, th<' overlying Lewisville member, which makes up from oO to 100 feel of the several hundred feet of total thickness of the Wood- bine, Ix'itig marine and so far as known lacking determinable fossil plants. The Woodbine sediments from a niaxinuim thickness of 300 to 500 foot in the Rod liivor • Am. Jour. 8H., 3d .lor., vol. 33, pp. MI-JMW, 1S87. ' Tail, J. A., Texan Oool. Sun-oy Kourlh Ann. Ropl., p.Z«, IWi. FLORA OF TIIK WOOPBIXE SAXH Al Ai; I 111 Its lil.lll-, TtX. 155 viillcv tliiii toward the soiitli or nro ropliiccd by clillVront litholobable and which has no basis of known facts." A thin! alternative which I regard as more probable than the two advanced by Stephenson is that the leaf-bearing sands of Red River which are referred to the lower Woodbine are the lime equivalent of what is called Eagle Ford in the Austin section. Either Woodbine time is represented there b}- the Buda limestone, which Stephenson states is highly improbable, or it is repre- sented bj- a break in sedimentation, which seems to me equally improbable, or it is repre- sented by the lower part of the Eagle Ford of that section. FAUNA. The fauna of the Woodbine, which comes from the upper part, or the Lcwisville marine member, is not extensive, but certain species are individually abundant at some localities. It comprises an almost unicjuc assemblage of shallow-water forms as identified by Cragin. including the following: Area galliennei var. tramatensis Cragin. Ostrea soleniscus Meek. Modiola filLsoulpta Cragin. Aguileria ciimminsi ^\Tiite. Cythorca Icvorotti <'ragin. Trisronarna siouxcnsis (Hall and Meek). Turritella renauxiana D'Orbigny. Cerithium tramatensis Cragin. Cerithium interlineatum Cragin. Pteria salint'n.sis Whito? Xatica humilLs Cragin. It contains also a considerable number of additional forms including uniinonites, not yet determined. It is unfortunate that no ade- quate study of the Woodbine fauna has ever been made. FLORA The flora of the Woodbine sand described in the following pages amounts to only 43 species. It is hence inadequate for a correct estimate « Stephenson, L. W., U.S. Geol. Survey Prof. Paper 120, p. 145, 1918. of its botanic facies. and the absence of forms normally jiresent in beds of e(|uivalenl age may, with much probability, be allributed to their lack of discovi-ry in the Woodbine, for as far as it goes it is a perfectly normal assem- blage of forms such as would be expected at this horizon. There are no known ferns or lower plants present, presumatdy because of the coarseness of the sediments and the trituration to which most of tlie plants have evidently been subjected. Only two gymnosperms are represented, a Podoznmittx and a Bnichi/phi/Ihnn. In this respect the Woodbine flora is more like the Dakota flora than it is like the corresponding floras of the Atlantic Coastal Plain, in which conifers are usually abundant individually and varied specifically. No monocotyledonous an- giosperms have been recognized in the Wood- bine, and this lack, like that of the ferns, is probably to be attributed to macerating and triturating water action. The dicotyledons, which comprise 41 of the 43 identified forms, represent 31 genera in 21 families and 15 orders. They are well .scat- tered among the families usually represented in Upper Cretaceous floras. Elements that are conspicuous b}' their absence when the Woodbine flora is compared with the Tusca- loosa flora of the eastern Gulf area, for ex- ample, are the numerous species of figs and magnolias, for there are only two of each of those types in the Woodbine, and one of the figs is a Dakota sandstone species and not a Coastal Plain species. Other notable absen- tees are Menispermites. Bauhinia, Lirioden- dropsis. Lefivminosites. CelastropJnjlhim, Gre- unopsis, Pterosper/mtfs. Sapotacitm, and other less significant genera. The genus Celas- trophyllum is especially abundant in the Tus- caloosa, in which 12 species have boon recog- nized, and it is almost equally abundant in the Raritan formation. No genus is represented in the Woodbine by more than two species except the form genus CarpnJithvs. The largest family is the Laura- ceae, with eight species; no other family has more than three species, and only two families, the Salicaccac and Magnoliacoae, reach that number. Similarly the largest order is the Thymi'loales, with eight species, and the only order that approaches it in size is the Ranalos, with five species. 156 SHORTER C0"NTRIBUnONS TO GENERAL OEOLOGY, 1921. The flora, as at present known, is inaiioqiiato to indicate the environment in which the plants lived, but it clearly contains no unusual fea- tures, and I can see no reasons for supposing that the physical conditions were different from those which have 1)een predicativl from the much lar<;or floras of the Tuscaloosa. Rari- tan. and Mathv formations, which have been discussed at length in various publications. A systematic list of the Woodbine Mom at Arthurs Blufi' follows: ( ivmnospermae: Podozamite,") lanceolatus (I.iiulloy ami Ilut- ton) F. Uraun. Brachyphylluin iiiaiTDcarpuiii fnriiKisiiin Bern'. Piootyledonae: M>Ticales: MjTieaceae: MjTica emarginata Ileer. MjTica lonpa i Hoi^n Ileer. Salicales: Salicaceae: Salix lesquereuxii Beiry. Salix (leleta Lesqiiereux? Populus harkeriana I.esquereux. I'rticales: Moraceae: Ficus daphnogenoifie.s (Ileerl Berry. FicuB glaacoeana I.esquereux. Platanalew: Platanaceae: Platanus latior (Leequereux) Kiiowltun. Ranales: Magnoliaceae: Magnolia specioea Heer. Magnolia lacoeana Les(iuereux. Lirio'lendron quercifolium Xewbern,-. Trochohno alabamen.si.« Berry, rinnamiiniuin newberrj-i Berry. Oinnainiiniiim membranaceum (Leequereux) Uolliik. Lauru.s plulonia Ileer. T/aurus antecedens I.esquereux. Laurophyllum minus Newberry. Myrlales: .Myrlaceao: Myrtxiniiim geinitzi CHeer). Uinbollales: ■Vraliaceao: .\ralia wellingtimiana I.esquereux. Aralia saportana Lesquereux. Cornaceae: Comophylluni vetustuin Newberry. Kricalea?: Andromeda novaecaesareae HoUick, Andromeda snowii I..esquereux. ICbenales: ICbenaceao: Dioapyros primaeva Heer. Rul)iales: Caprifoliaccae: Viburnum robustum Lesquereux. Position uncertain: Tricalycites papyracetis Newberrj'. CarpolithuH sp. 1. Carpolitluis sp. 2. Caipf)lithus sp. :?. AGE OF THK FLORA. RELATION TO THE FLORAS OF OTHER UPPER CRETACEOUS FORMATIONS OF THE COASTAL PLAIN. fuliilion to the /iiiif/(ii sand jhini. — The geo- griipliicMlly nearest flora to that of the Wood- bine sand is that Iniiiid in llic Biiigcn sand of ArUansas. The Bingen is generally recogni/i'd as being, in |>arl, al least, tiie e(iuivalenl of the Wooilliine and also as re|)re.seiiting all the U|)per Cretaceous of .VrUansas below the finxji/ni i>iiii(/i nimi /.one. The latter fact is of especial importance in the final determination of the age of the Woodhine, for if llie Bingen in its eastcriiMiost exposures represents the whole of the int<'rval lepre.sented in northern Te.\«is bv the Woodbine and lOaglc Ford formations tliere is no apparent reason why a part of the FLORA OF THE WOOPBIXF. SAXn AT ARTHURS BLUFF, TKX. 157 Woodbini" may not also rcpn-^i'iil tlu' Kaglc Ford. Such a conclusion would he in accord xvith the evidence of the fossil j)lants. The flora of the Bingen sand is known oidy from a small collection made by H. D. Miser in Pike and Howard counties, Ark.' It com- prises but 27 named forms, and these are not all from one level: hence comparisons between it and that of the Woodbine are limited. In spite of this the Woodbine contains nine species that are common to the Bingen sand, including the Dnralquea, which is common in the upper Bingen and is confined to these two formations. The upper Bingen was considered by me as the e(iuivalent of the upper part of the Tusca- loosa and the Eutaw formation of the eastern Gulf area, and the lower Bingen as the equiva- lent of the lower Tuscaloosa and the Raritan formation. Unfortunately for ease of correlation and in- telligible discussion all the I'pper Cretaceous formations of the Coastal Plain have been based up^mlithologic differences instead of upon their contained famias and floras, iuid as they over- lap iuid inlergrade laterally the limits of the same formation are not chronologically equiva- lent from locality to locality, so that precision in correlation must await the discover}- and study of much more extensive paleontologic materials than are available at the present time. Although the Bingen has been separated by Miser into upper and lower members, the known flora was found near the top of the lower mem- ber and near the base of the upper member and hence could not be expected to be as deci- sive as if it represented both earlier and later Bingen time. The present Woodbine flora comes from a single horizon and locahty, and what the other 300 to oOO feet of the Woodbine would show if the flora were knowTi in its entirety can only be surmised. Hence, inas- much as names must be used m any discussion, it must be borne in mind that when I speak of the Wdbine plants in other fornuilions. tliat Arthurs Hlufl" is approximately on the bound- ary between the lower and upper members of the Bingen as rect>gnized by Miser in Arkansas in the specific area where he collected tlie fossil phmts. Relation to TuKcaJonmt flora. — The flora of the Tusctdoosii formation is extensive, com- prising 151 describeil species, recently mono- graphed.'" The Tuscaloosa occupies the same stratigraphic position with respect to the Eutaw formation of the eastern Clulf area that the Woodbine tloes with respect to the Eagle Ford formation of the western (lulf area, and both the Eutaw and the Eagle Ford contain com- parable marine faunal elements. The Tusca- loosa formation has been shown to be progres- sively younger when traced northward from western central Alabama, and in the report just cited 1 have showni its delta character and probable clu'onologic equivalence with a part of the marine Eutaw formation. I suspect that the Woodbine might also be uiterpreted as made up of continental, delta, and marginal deposits, with sinxilar relations to the marine Eagle Ford, but I have no basis for this inference except the writings of others. This would afford an excellent subject for field study. The Woodbine and Tuscaloosa floras have 22 species in conunon, so that it seems clear that the Woodbine and Tuscaloosa for- mations are equivalent, at least in part. Whether the Tuscaloosa elements that are con- spicuously absent in the Woodbine represent real or only apparent differences can not be determined. I incline to the opinion that these differences are ordy apparent. Relation to floras! of other formations of the Coastal Plain. — The relation or degree of resem- blance between the Woodbine flora and that of geographictdly more remote formations of the Atlantic Coastal Plain is well shown in the accompanying table of distribution. The Woodbine contains 20 species common to the Raritan, IS common to the Magothy of the New Jersey-Maryland region, and 2.5 common to the two combined, thus emphasizing a well- known floral similarity seen throughout the Coastal Plain. This ma}- mean that the Wood- bine is equivalent to the upper Raritan and the Magothy, or simply that it is equivalent to the " Berry, K. W., U. S. tipol. Survey Prof. Paper in. 1919. 158 SHORTER CONTRIBtTTIONS TO GENERAL GEOLOGY. li>21. Miigothv, as a great many upper Raritau spe- cies survived into Magothv time. There are nine species common to the Woodbine and Bhick Creek formations, but six of these are widespread and king-lived forms without espe- ciid significance. Five of the Woodbine species are common to the Eutaw flora, and one ex- tentls Its liigh as the Riidey formation of the eivstern Gulf area. RELATION TO THE FLORA OF THE DAKOTA SANDSTONE. The llora of the Dakota sandstone is very large, embracing over 400 described species. It is also very obviously not ail of the same age, but outside of certain areas in Kansas and Ne- braska no data are available for determining what part of Dakota flora came from beds that merit that term and what part came from beds sine* discovered or suspected to be different. Of the 4.3 Woodbine plants .30 arc species of the Dakota sandstone flora. The community of facics is thus very great, and it is significant that of these 30 common species all but 10 arc forms which their range in other formations proves to represent what for want of a better term might be termed the true Dakota flora, and tiiree of these 10 are only doubtfully deter- mined from the Morrison formation in Colo- rado " and these three occur definitely in the Tuscaloosa, Raritan, and Magothy formations, so that really 2-3 of the Woodbine plants are true Dakota forms. I regard ttie Woodl)in(> as syn- chronous with this undelimited Dakota sand- stone, which I regard as bearing the same rela- tion to the Benton as the Woodbine does to the Eajrle Ford and the Tuscaloosa docs to the Eutaw. RELATION TO THE FLORA OF THE CHEYENNE SANDSTONE. The relations ix-tween the Woodbine flora and that of the Mentor formation of central Kansas are unknown anrl will remain so imtil the Mentor flora is carefully collected and studied. The Cheyenne sandstone of soul linn Kansas contains a considerable flora, re<<'ntly ^ studied by me, but there is not a single species common to the Woodbine and Cheyenne. Al- though both floras consist largely of so-called Dakota forms, some of these in the Cheyenne " KnowUcin, K. H., Am. Jour. Hcl., Itbscr., vol. «, pp. 189-lM, 1020. sandstone are forms that may have come origi- nally from the imrecognized Cheyenne sand- stone or the supposed <'quivalcnt Mentor forma- tion, and not from the true Dakota sandstone. The majority lack an outside distribution. The Woodbine plants, on the other hand, as shown by the comparisons in the preceding sections, are nearly all well-known species of formations of known age of the Coastal Plain. There can not be the slightest doubt that the Woodbine sand is younger than the Chej-enne sandstone and synchronous with the true Dakota sandstone of the Western Interior. The Woodbine is also certainly younger than those beds in the West formerly confused with the Dakota and now Iviiown as the Purgatoiro formation. RELATION OF THE WOODBINE FLORA TO FLORAS OF OTHER AREAS. None of the Woodl)ine species have been identified in the formations of the Montana group. Eight of the Woodbine species are fountl in the Atane beds and six in the Patoot beds of western Greenland. The similarities of the flora to European Upper Cretaceous floras are reasonably close, but these similarities naturally seldom extend to the identical species, and where they do the particular forms are wide-ranging and long- lived species of slight value in precise correla- tion, as might be expected. The Woodbine contains five species common to European beds referred In (he Cenomanian and three addi- tional that are tentatively recognized at this European horizon. .Similarly two species are identical with forms that occur in the European Turonian. None of the species are known in ihv Emscherian of I']uroj)e, although I)fir(il(/iie(i ijisu/niformis of the Woodi>ine is, as its name indicates, very close to the European .Senonian species Dewalquen iiisif/nift Debey. The intrinsic character of the Woodiiine flora necessitates considering it as either Ceno- manian or Turonian wlicn judged according In Euro[)ean standards. From its relationshij) with other more extensivi^ fonnations of the Coastal Plain such as the Tuscaloosa and Ma- gothy, in which the evidenc(> is clearer, I Wdiiid be incliiM'd to considi-r (lie Woodbine flora as of Turonian age FLORA OF THE WOODBINK SANP AT AIlTHURS BLUFF, TKX. Outside distribution of Woodbine flora. loll Dakota T"*^ ■^"-j'uor Bincen saiid. Rari- taii forma- tion. Mag- oihy forma- tion. Bliick I'reok forma- tion. Eiitaw forma- tion. Rlploy forma- tion. Atano bads. Pa toot b«ds. Horri- Ron fornia- lioii. Ceno- man- lan. Turo- nlaii. Podozamiles lanceolatus X X X X X X Bracl^y^)h^•lhlm macrocar- l>uin lormosum X X ? M \riia emargiuata X X X X X x' X X "x" X X Mvrita longa X X X X Salix lesqucreiixii X X •> X Salix di-k'ta? Popviliis harkeriana X X Kcus da|)hnogenoide8 X X X ? Fious filasi'ooana Platanus latior .Nhvfinulia spociosa X X X X X X •yl-^- .Maf;iiolia lacoeana Liriodoiulron miorcil'olium . . Trichodendroiaea rhomboid- ens X Dewalquea insigniformiB X Pahu'ocassia laiirinea X X X .^ Colutoa primordialis .\ X X X X X Sapindus morrisoni X X Rhus redditilormis Rhamnus teiiax X X X X ? Ziz\-phu3 lamarensis Cissites forniosus X X X X X X Sterculia higiibris Benzoin veniistum Malapoenna falcifolia Oreodaphne alabamonsis. . . . X X X X CHnnamomum newberryi Cinnamomum membrana- ceum X X X X X X X X X X X X X X X ? Launis plutonia X X X X X X ? Lauras antecedens? I.aiirnphvllum minus?. X X X M\Ttonium geinitzi X X X X X X X X X Aralia wellingtoniana Aralia saportana? ConiophvUum vetustum .... X X X X X Andromcila novae-caesareae. X X Andromeda snowii X X X Diosp\Tos primaeva X X X X X X X X Viburnum robustum Tricalycites papjTaceus X X V 1 ! ■■ SYSTEMATIC ACCOUNT OF THE FLORA. Phylum CYCADOPHYTA. Genus PODOZAMITES F. Braun. Podozamites lanceolatus ( Lindley and Hutton i F. Braun. Plate XXXVI, figure 2. Podozamites lanceolatus (Lindley and Hutton) F. liraiin, in MUnster, Beitriige zur Petrefactenkunde, vol. 2, pt. 6, p. 53, \9AZ. Dawson, Roy. Soc. Canada Trarw., vol. 3, sec. 4^ p. 6, pi. 1, fig. 3, 1886. Lesquereux, U. S. Geol. Survey Mon. 17, p. 28, pi. 1, figs. 5, 6, 1892. Newberry. U. S, Geol. Survey Mon. 26, p. 44, pi. 13, fig. 2 [not figs. 1, 3, 4], 1896. Penhallow, Canada Geol. Survey Summar\- Kept. 1904, p. 9, [1905J. Fontaine, in Ward, U. S. Geol. Survey Mon. 48, p. no, pi. 24, figs. 17-20, 1905; U. S. Get.1. Survey Twentieth Ann. Rept., pt. 2, p. 360, pi. 63, fig. 4; pi. 66, fig. 4; pi. 67, figs. 3, 4, 1900. Knowlton, Smithsonian Misc. Coll., vol. 50, p. 120, 1907; U. S. Geol. Survey Prof. Paper 85, p. 52, 1914. HoUick, U. S. Geol. Survey Mon. 50, p. 35, pi. 2, fig. 1, 1907; New York Bot. Garden Bull., vol. 8, p. 155, pi. 162; pi. 163, figs. 2, 3, 1912. Berrj-, Torrey Bot. Club Hull., vol. 39, p. 391, 1913; Maryland Geol. Survey, Lower Cretaceous, p. 341, pi. 53, figs. 5, 6, 1911; Tcrrey Bot. Club Bull., vol. 38, p. 410, 1911; Maryland Geol. Survey, Upper Cretaceous, p. 772, 1916. Pinnis distantibus, alternis oppositisve, elongatis, basi sensimangu.'itati.s, inferioribus lanceolato-linearibus, supe- rioribus elongato-ellipticis; nervis crebria. — Schimper, 1870. 160 ;H()RTKR COXTRIBrTIONS TO GENEKAL C.EOLOGY, l'J21. Tliis is 11 species of great geologic range, heiiig recorded from the Jurassic to the Upper Cretaceous. The geogiaphic range is equally extensive, embracing two continents, North Ajncrica and Europe. It is ([uite probable tliiit the species is composite, luit no certain grounds for segregation are apiJiuout. Some students may doubt tlie wisom of cor- relating both Lower and Upper Cretaceous fornis with a species which is essentially a Jurassic type, but specific differentiation fouiul- ed merely upon stratigraphy has gone astray so often that in cases Uke the present s3-n- thesis may well precede analysis, and it might be added that this was the view taken l)y IIol- lick" with reference to material from Glen Cove, Long Island, suid by ^'elenovsky " in studying the Cenomanian flora of Bohemia. Forms indistiiiguishable from the type of this species occur in both the I'atuxent and Patapsco formations of tlie Potomac group, as well as in the Kootenai, Dakota, Black Creek, Karitan, and Magothy formations. Phylum CONIFEEOPHYTA. Genus BRACHYPHYLLUM Brongniart. Brachyphyllum macrocarpum formosum Berry. Plate XXXVI, fisiirol. Brachyphyllum macrocarpum Berry, Torrey Bot. Chil) Bull., vol. 38, p. 183, 1910 mot Newberry, 1896t; vol. 38, p. 420, 1911. Brachyphyllum macrorarpum formosum Berry, iilem, vol. 39, p. 392, pi. 30, 1912; V. S. Geol. Survey Prof. Paper 84, p. 106, 1904; Prof. Paper 112, p. 59, pi. r,, fig. 9, 1919. Slender elongated twigs, pinnatdy liranched, covered with medium-sized crowded, appresseil leaves, spirally arranged. Leaves hlunlly pointetl, smooth, thick. In the consideration of the various specimens whicli have been referred to Brachyj)ln/llum miwroairfium, a very considerable variation within certain fixed limits is at once obvious. This variation is usually one of size, the more slender specimens being at the same time smoother. This characteristic has been fre- quently noted tiy ine and is commented upon in |jrint l)y Knowlton," who in discussing the II BoUlck, Arthur, V. S. Oeol. Survey Man. Ml, p. 3S, ISOS. " Velnnoviky, Jrawt, Die Oymnmiiwineii dor Wlhmlsrhoii Kriililo- rormntlnli, p. II, pi. 2, (Igs. II 10, 21. IKV,. '• Knowlton, K. U., U.S. Oeol. Survey null. 10,1, p. 2V, pi. 4, Unn.U.n, l«00. younger forms from Wyoming suggests that the specieson the vergeof extinction becamesmaller in its proportions. In studying the material from the South Atlantic and Gulf .States a con- stant ditrcreiice in size was noticed. This may reflect a slight difference in < liniatic condi- tions, and all the forms may be interpreted as the variations of a single species -in fact, the specimen from the Raritan formation in New Jersey illustrated in Newberry's (igure 7"^ is approximately of the same size a.s the forms from the Montana group of the West and is associated with the normal stout, club- shaped type. That the variety ha.s no par- ticular stratigraphic significance is indicated by its abundance at a horizon as low as the ba.sal part of the Tuscaloosa of Alabiuna and its presence in tlie Woodbine sand of Lamar County, Tex. In general the present variety occurs in later and more southern beds than the type, a difference which might be ascribed to the fact that only the slender terminal twigs are preserved. This explanation is regarded as improbable, however, for the same reasoning should hold good for the areas where onl}' thicker twigs have been found. The remains are usually much macerated and broken, and the immediate cause for the recognition of a new variety was the discovery of a relatively large specimen from the Magothy formation of Maryland, which showed such striking unlikeness to the typo that separation was demanded and specific differentiation was even considered. In view, however, of the occurrence of both forms in assiM'iation in Maryland and tlie well-known variation not only of the type but of coniferous foliage in general, it seemeil wiser to consider tiie present form as a variety of the type, whicli as time ])rogressed supplanted it to a large extiMil if not altogether. Tlu^ new specimen from Maryland showed the lerminal jiait of two approximately parallel and curved lwigsal)out Vl centiinolors in length, united proximad. Those in their largest portion ai(< only 11 millimeters in diame- ter. At intiM-vals of ;{ to .") millimeters sub- opposite lateral branches arti given off in a pinnate manner. Tlu^se are relatively much elongated, curved, and slender, averaging about 4 centimotei-s in length by 2 niillimetere ■ Nawberry, J. 8., U. 8. Oeol. Survey Mon. 24, pi. 7, nn.i. 1-7, IKiM. FLORA OF Till-: WOOHBIXK SAND AT AHTIiritS BUTF, TKX. ir.i in (liniiu'tcr. hluiuly puintod. ami not tu])t'riiii; to any appri'ciahlc cxti-nt. A few of tlicsc lateral branches fork psoiulodichotomously. and some of them jrive off toward thi>ir distal ends tiny lateral hranchlets less than 1 centi- meter in len<;;th and about 1 millimeter in diameter. The general proportions are thus decidedly diflerent from those of the supposed parent t_vpe. The leaves are slifjhtly smaller and smoother and relatively somewhat more elon- gated, at the same time laekinjj the apical papilla and the convergent striae. The form is much more graceful in api)earance than the type and in its general aspect suggests the Lower Cretaceous genus Arihrofaxopsis of Fontaine. \Mule tiny species of Brachyphyllum like Brachiipiiyllum microcladnm Saporta, of the Xeo-Jurassic, have been described, the new variet}- is even more slender than Brachy- phyllum yracUf Brongniart. of the Jurassic. The most closely allied form known appears to be one from tiie .Vlbian of Buarcos, in Por- tugal, described by Saporta '" as Brachy- phylbim ohesiforme elongatinn. The present form also shows considerable resemblance to BrachyphyUuni crassicaiilc Fontaine, of the Patapsco formation in Maryland and Virginia. This variety is abundant throughout the Tuscaloosa formation and in the basal part of the Eutaw formation in Alabama and west- ern Georgia and occurs also in the Woodbine sand of Texas but is kno\vn in Maryland only from a single locality. It is also confincid to a single locality in North Carolina, where it is not at all uncommon but is not especially well preserved. Phylum ANGIOSPERMOPHYTA. Class DICOTYLEDONAE. Order MYKICALES. Family MYKICACEAE. Genus MYRICA De CandoUe. Myrica emarginata Heer. Mi/rka emarginata Heer, Flora fossilis arctica, vol. (i, Abt. 2, p. 66, pi. 41, fig. 2, 1882. I.eaquereux, The flora of the Dakota group, p. 67, pi. 12, fig. 1, 1892. Xowterry, The flora of the AmUoy ila>-s, p. 62, \A. 41, figs. 10, 11, 1896. Berry, New Jersey Geol. Sur\-ey Bull. 3, j). 104, jil. 10, fig. 5, 1911; U. S. Geol. Survey Prof. Paper 112, p. 73, pi. 13, fig. 4, 1920. " Saporta, Gaston de, Contributions k la flore fossile du Portugal, p. 176, pi. 31, fig. 14, 1894. Ileer's (lescri[)tion, published in 1SS2, is as follows: M. foliis olilongis, intogerrimi», apioe i>umn;inuii.-i, l)ii.»i attonuatiH, norvis sei-umlariis sulUilissiiniH. Tliis species, which was described from speci- mens collected in the Atane beds of (Ireenland, has been recorded from the Karitan ami Tusca- loosa formations of the Atlantic Coastal Plain and from the Dakota sandstone of the Western Interior. It is somewhat variable in form but may be characterized as follows: Leaves obovate, widest at the rounded, truncate, and more or less eraarginate ape.\, with entire margins narrowing to the cuneate base. Midril> mediiimly stout. Secondaries, five thin pairs, subo[)posite, diverging from the midrib at angles of about 4.5°, campto- drome. The reference of this species to the genus Myrica is entirely problematic. It is astonish- ingly close to a form from Niederschoena, Saxony, described by Engelhardt '" as Minm- sopx ballotaevides. Myrica longa (Heer) Heer. Plate XXX IX, figure 5. I'roUoidejs loiigus Ileer, Flora fossilis arctica, vol. 3, Al>t. 2, p. 110, pi. 29, fig. 8b; pi. 31, figs. 4, 5, 1874. Dawson, Roy. Soc. Canafla Trans., vol. 1, sec. 4, p. 22, pi. 2, fig. 8, 1883. Fri<', Archiv naturw. I.aniles. Bohmen, vol. 4, Xo. 1, pp. 18, 94, 1878. Myrica longa (Heer) Heer, Flora to8.silis arctica, vol. 6, Al)t."2, p. 65, pi. 18, fig. 91); pi. 29, figs. 15-17; pi. 33, fig. 10; pi. 41, fig. 4d, 1882; vol. 7, p. 21, 1883. I.esquereux, U. .S. Geol. Survey Mon. 17, p. 67, pi. 3, figs. 1-6, 1892. Bartsch, Iowa Univ. Lab. Nat. Hist. Bull., vol. 3, p. 180, 1896. Knowlton, V. S. Geol. Survey Twenty-first Ann. Kept., pt. 7, p. 314, pi. 39, fig. 7, 1901. Berry, Torrey Bot. Club Bull., vol. .33, p. 170, 1906; Maryland Geol. Survey, Upper Cretaceous, p. 812, pi. 57, figs. 1-3, 1916; Torrey Bot. Club Bull., vol. 44, p. 175, 1917; U. S. Geol. Survey Prof. Paper 112, p. 74, 1919. Leaves of various sizes, linear to lanceolate, with a stout midrib, numerous thin, ascending, camptodrome secondaries, entue margins, ob- tuse!}' pointed apex, naiTowly decurrent base, and long, stout petiole. This species, which Heer described as a ProUoidcs and siibsef|uently transferred to the genus Myrica, has a particularly wide " Engelhardt, Hermann, Natiirwis.s. licsoll. Isis in Dresden Abli. 7. Jahrg. 1891, p. 9s, pi. L', (Ic 13, 1S92. 162 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY, 1921. distribution. It occurs in both the Atiine and Patoot beds of wostorn Greonland, in the Da- kota sandstone of the West, in the Majjothy formation of Marvhmd, and in the Bingen sand of .Vrkansas and is very conmion in the Tuscaloosa formation in western central .Vla- bama. Abroad it has been recorded from the lower Turonian of Bohemia. Dawson recorded it frona Peace River m Northwest Territory. It occurs in the Dakota samlstone in Wood- bury County, Iowa, in Ellsworth County, Kans., and near Lander, Wyo. It is thus a member of the true Dakota sandstone llora. This is also emphasized by its distribution in the Upper Cretacwius of the Atlantic Coastal Plain. Although abundant this species lacks good diagnostic characters, and its botanical affinity is therefore uncertain. Its relation to Myrica is o.xtremely doubtful. Order SALICALES. Family SALICACEAE. Genus SALIX Liimg. Salix lesquereuxii Berry. Salix Uiquereurii Berry, Torrey Bot. Club Bull., vol. 36, p. 2.i2, 1909; vol. 37, pp. 21, 194, 1910; New Jersey Geol. Survey Bull. 3, p. 114, 1911; U. S. Geol. Survey Prof. Paper 84, pp. 33, 109, pi. 7, figs. 11-13, 1914; .Marj'land Geol. Survey, Upper Cretaceous. p. 814, pi. 58, figs. &-8, 1916; Torrey Bot. Club Bull., vol. 44, p. 176, 1917; U. S. Geol. Survey Prof. Paper 112, p. 76, 1919. Salix proteiufolia Leequereux, Am. Jour. Sci.. 2d ser., vol. 46. p. 94, 1868 (not Forbes); The Cretaceous flora, p. CO, pi. 5, figs. 1-4, 1874; The Cretaceous and Tertiary floras, p. 42, pi. 1, figs. 14-16, pi. 16. fig. 3, 1883; The flora of the Dakota group, ]). 49, 1892. Newberry, The flora of the Amboy clays, p. 66, pi. 18, figs. 3, 4, 1896. Kurtz, Mu8. La Plata Rev., vol. 10, p. jl, 1902. Berry, New Jersey Geol. Survey Ann. Kept, for 190.'), p. 1:59, 1906; Torrey Bot. Club Bull., vol. 33. p. 171, pi. 7, fig. 2, 1906; Johns Hopkins Univ. ('ire, new ser., No. 7, p. 81, 1907. Salix proUae/olia Innrfifolia Lcs<|uereux, The flora 'B, p. 72 (part), pi. 32, fig. 11, 1896. l)fualfjurii gromlandirn Heer. Newberrj', The flora of the Amlioy clays, p. 129 (part), pi. 41, fig. 12, 1896. Leaves ovate-liin<'eoIate, somewhat more Miiifiiuiiiii' idiove than below, \iirinble ill size. ranging from 6 to 12 centimeters in length and from 1.1 to 2.2 centimeters in greatest width, which is usually slightly t)elow the middle. Petiole stout, much longer than in Salix Jieiitosa, having a ma.ximiim length of 1.2 centimeters. Midrib stout below, ta])ering above. Secondaries numerous, in some speci- mens as many as 20 pairs; they branch from the midrib at angles of about 45° and are sub- parallel and camptodrome. This species is e.xcecdingly variable, as might be expected in a Salix, and Lesciuereu.x estab- lished several varieties, of which at least one, Umarifolia, is referable to Salix flexuoaa New- berry. Some of Lesqucreu.x's forms are dis- tijiguishable with difliculty from Salix fltmosa, and this is especially shown in the leaves which he figures on Plate I of his "Cretaceous and Tertiary floras." They are, however, larger and somewhat more robust, of a thicker texture anil broadest near the base, from which they taper upward to an exceedingly acuminate tip. In general Salix lesquereuxii is a relatively much broader, more ovate form, with more numerous and better seen secondaries and a longer petiole. This species is an exceedingly abimdimt Upper Cretaceous type in both the East and the West, ranging chronologically in the Coastal Plam from the base of the Raritau formation to the top of the Tuscaloosa formation, and possil)ly through the Eutaw formation as well. It is al)undiuit in the Afagothy and Black Creek formations, mcludijig the Middendorf arkose member of the Black Creek. It is widespread from the bottom to the top of the Tuscaloosa formation in Alabama imd occurs in the Bingen sand of Arkansas. In the West it is a member of the Dakota sandstone ilora and is not |)resent in the older Cheyenne sandst*>ne flora, nor is it known from ])eds younger thiui the Dakota sandstone in that region. It is one of the forms recorded by Kurtz from (he Upi)er Cretaceous of Argentina, indicating, if the identification is correct, wliicb is doubtful, a very consider- able migration during early Ip per Cretaceous time, Kecentl}^ Know lion'" lias tentatively identi- fied this species from (lie type section of ilie Morrison formation in Colorado, ' KiKiwItnii, V. H., Aio. Jour, Sol., Ith M21. the genus F(ci/.«, where llu-v properly Ix'long, they find their iiinnity in the group which incUiiles, among others, such fossil species as Fictts elongata Hosius, Ficus berthoudi Les- (piereux, Ficus suspecla Velenovsky, and Ficus kransiana Heer. This species has been found to be variahle in size, nuiging in length from 1 1 to 22 centimeters and In width from 1.9 to 3.7 centimeters. It is usually widest in the lower half of the leaf, although In some specimens the base is narrow and the wiilest part is toward the middle. In all unequivocal material the upper half of the leaf is narrow and is produced as a long, slender, commonly recurved tip, which is one of the characteristic features of the species. This tip is strictly comparable with the "dripping points" developed on various modern leaves in the regions where precipitation is heavy. Ficus (laphnoijfnoidff! is a widespread and common form, ranging from Marthas \'ineyard to Alabama, Arkansas, and Texas in the Atlantic Coastal Plain and from the Northwest Territory to Kansas and Nebraska in the Western Interior region. It is a member of the Dakota sandstone flora and does not occur in the older Che^ejuie sandstone so far as my observations go. It was reported by Ward from the Cheyemie sand- stone at Chatman Creek, Kans., but Ward's material, which I have before me, is not this spe- cies but a leaflet of Sapindopsis. Kiiowlton'" has recently tentatively identified this form from the type section of the Morristm formation at Morrison, Colo. The species is not uncom- mon in the Woodbine sand at Arthurs Bluff, Tex. Ficus glascoeana Lesquereux. Ficua glascoeana Lc-sqiicri'ux, U. S. (icol. Survey Terr. Rept., vol. 8 (CretaceoUH and Tertiary (lorawi, p. 4S, 1883; U.S. Geol. Survey Mun. 17, p. 7(1, \>\. 13, figs. 1,2, 1891 [1892]. Knowlton, U. S.Geol. Survey Twenty-first Ann. Rept., pt. 7, p. 317, \'.m. IjCaves large, ohlong-ovalc iji general outlijU", with an obtusely pointed apex and a broadly rounded or cuneate, ultimately slightly de- current base. Leugth from IS to 20 cenli- meters; maximum width, at or below the middle, 6. .5 to 7.5 centimeters. Margins entire. Texture coriaceous. Surface polished. Peti- » Knowllon, F. H., Am. Jour. Bel., 4tb Mr., vol. «, p. ign, igjo. ole missing. Midril) very stout and ])romi- nent on the under surface of the leaf. Secon- daries numerous, thin, diverging from the midrib at wide tuigles, suijparallel, straight, ascending, joijiijig one another by abru|)t curves subparallel with anil close to the nuir- gms. This species, which is of the same generiJ type as Ficux ntdviiia Ileer,-" was described by Lesquereux from material collected 2 J miles soutli of filascoe, Kans. .So far as known it is cojilined to the Dakota sandstone juid to the Woodbine sand at Arthurs Bhifl', Tex., and the remains were usually much broken before fossilization. Ficun atavina, which is closely related to it, was also a stiff form, usually found in a broken condition. It has an ex- tensive range, occurrijig in the Atane antl Patoot beds of Greenlimd, the Magothy for- mati(m and the Middeiidorf arkose member of the Black Creek formation in the Atlantic Coastal Plam, the Turoniim of Bohemia, antl the Gosau betls of Tyrol. It seems very prob- able that Ficu.f (jlascoeana is genetically re- lated to the more widely distributed Ficus atavina. Order PLATANALES. FamUy PLATANACEAE. Genus PLATANTIS LinnS. Platanus latior (Lpsqucreuxi Knowllon. I'lalanus arcroidi^'! Goeppert var. lalior Losqucroux. .\m. ,Tour. Sci., 2(1 ser., vol. 4G, p. 97, 1808. Platanus lalior (l.«8quereux) Knowlton, V. S. Gool. Sur- vey Bull. 152, p. 170. 1S98: U. S. Geol. Survey Twenty-first Ann. Rept.. pt. 7, p. 314. 1901. norry, V. S. Geol. Survey Prof. Taper 112. p. 84. I!t20. Platanus pnmacva Lescpiereux. The Cretaceous llora. ]). 09, pi. 7, fig. 2;pl.2fi. (ig. 2, 1H74; The lloraoftlie DakoU group, p. 72, pi. 8, figs. 7, Sir. pi. 10. fig. 1, 1892. Tvcaves large, palmately trilobate, broadly riiomboidal. Lenglli about 17 centimeters; maxiinniii wicjili iiliDiil 1 ."i I'ciil iinclcrs. Mar- gins somewhat iiTcguliiily dentate, entire at the biDadly cuneate itasc. Lateral lobes short; intervening sinuses scarcely differentiated. Petiole long mill stout. Priiiuiries stout, three in luimher, diverging at it ih'mi- liie l)ase in the material from Alabamn Iml idiiinuuily suprii- basilar in the forms from the |)aUota sandstone. \'enalioii strictly platanoid. Textine coria- ceous. • npcr, 0»wftlil, Flnrii fiwllls nrrlli-n, vol. fl, Abt. 2, p. 09. pi. II, fln. Sb.Tb.Hb: pi. 17.nf.sb; i.l.lli.tlir.lh; |.|.a).ng<.1.2,l»a. ll.oitA OF Tin; WIXIPBIXE SAXD AT AliTlirjiS IlI.rFF, TEX. 165 This iiiic largo spinies is very nhmulaiit in tlie Dakota sandstone of Kansas. Ncliraska, ami Minnesota, and Lesqucrcux difTcrentiatcd three varieties, iniegrifolia, subintegrifolia, and gran- did en tata. Order RANALES. Family MAGNOLIACEAE. Genus MAGNOLIA Linnfi. Magnolia speciosa Heer. Plate XL, figure 6. Magnolia speciosa Heer, AUjr. schweiz. Gesoll. Resammt. Xafurwiss. Bern Xeue Dcnskchr., Hand 23, p. 20, pi. 6, tig. 1; pi. 9, fig. 2; pi. 10, fig. 1, 1809. I.eequereux, The ("retareous and Tertiary floras, \^. 72, 1S74; The flora of the Dakota group, p. 202, pi. 60, figs. 3, 4, 1S92. IToIlick, Xew York .\cad. .''ci. Trans., vol. 12, p. 234, pi. 7, fig. 4, 1893; Torrey Bot. ("lub Bull., vol. 21, p. 60, pi. 178, fig. 5, 1894; Geol. Sor. .Vmerira Hull., vol. 7, p. 13, 1895; The Cretaceous flora of soulhorn Xew York and New England, p. M, pi. 19, figs. 1—1, 1900. Knowlton, V. S. Geol. Survey Twenty-first .\nn. Kept., pt. 7, p. 318, 1901. Smith, On the geology of the Coastal Plain of Alabama, p. 348, 1894. Berry, Torrey Bot. Club Hull., vol. 31, p. 7G, pi. 3, fig. 10, 1004; vol, 32, p. 4(1, pi. 2, figs. 4, 5, 190.5; XewJerseyGeol. Survey Bull. 3, p. 129, pi. 14, fig. 3, 1911; Torrey Bot. Club Bull., vol. 39, p. 395, 1912; U. S. Geol. Survey Prof. Paper 112, p. 8S, pi. 18, figs. 3, 4, 1919. Magnolia auriculata Xewberrj', The flora of the Amboy clays, p. 75 (parti, pi. 41, fig. 13; pi. 58, fig. 10, 1896. Heer's description, published in 1869, is as follows : M. foliis maximis, coriaceis, o\ato-ellipticis, apice longe attenuatis, valde acuminatis, basi in j)etioluin validum attenuatis, nervo primario cra.sso, ner\is secundariis valde curvatis, camptodromis. This species is somewhat variable in size. The American material, which is somewhat smaller than the type material from Moletein, Moravia, ranges in length from 8. .5 to 19 centi- meters and in ma.ximum width from 4 to 7.5 centimeters. It is ovate-elliptical, with the apex more or less produced and the base de- current. The midrib and petiole are stout. The secondaries arc well marked, campto- drome; they number seven to nine pairs and are subopposite, branching from the midrib at angles of about 1.')° and curving ui)wiinl. Tiie texture is coriaceous. This species, which was de.scribeil originally from specimens collected in tlie Cenomanian of Moravia, has a wide range in .\merica. Typical leaves occur in the Dakota sandstime. It is present on Marthas \incyard and Long Island and in the Raritan and Magothy formations of New Jersey. It is present at .Vrthurs Bluff, Tex., and was recorded by Knowlton from the Woodbine at Rhamey Hill, Dcuison, Tex. Magnolia lacoeanu Lesquercux. Magnolia lacomim T.esquoreux, The flora of the Dakota group, p. 201, pi, (iO, fig. 1, 1892. Newl>erry, The flora of the .\mboy clavs, ]i. 73, pi. .">5, figs. 1, 2, lS9(i. HoIIick, The Cretaceous flora of southern Xew York and Xew England, p. (15, ])1. 17, fig. 2, 1901). Berry, Torrey Hot. Club Hull., vol. 37, p. 23, 1910; New Jersey Geol. Survey Bull. 3, p. 134, pi. 16, fig. 2, 1911; Maryland Geol. Survey, Upper Cre- taceous, I). 832, pi. 70, figs. 1, 2, 1916; U. S. Geol. Survey Prof. Paper 112, p. 91, pi. 17, fig. 9, 1919. Leaves broadly oval to almost orbicular, obtuse or abruptly pointed above and rounded to a somewhat cuneate base below, 10 to 12 centimeters in length bj^ 8.5 to 9.5 centimeters in maximum width. Midrib stout, somewhat flexuous. Seccmdaries numerous, campto- drome, rather stout, 10 to 12 pairs; they branch from the midrib at acute angles, imme- diately curving outward, forming festoons near the margin, which is somewhat undulate in one specimen that Xcwberrj- referred to this species. This species differs from its contemporaries, especially in its nearl\- round outline. Les- quereux finds a resemblance to Magnolia inghfieldi Heer, from Greenland, and it also suggests some of the Arctic forms which have been referred to Magnolia capellinii Heer. Although this s[)ocies is reported from points so widely separated as Marthas Vineyard and Kansas, it is nowhere abundant and is usually poorly preserved, suggesting that the leaves were readily macerated. It also occurs in the Magothy formation of New Jersey and Mary- land and the Tuscaloosa formation of Alabama. The tyi'e locality was Ellsworth Count}^ Kans., and the range elsewhere of this species would indicate that this locality was in the true Dakota sandstone. 166 SHORTER CONmiBrTIOXS TO GKNERAI, GEOI.onY, 1021. Genus LIRIODENDRON Linn«. Liriodendron quercifolium Newberry. Plate XXXVI, figure S. Liriodendron quercifolium Newberry, Torrey Bot. Club Bull., vol. 14, p. 6, pi. 02, fig! 1, 1887; U. S. Geol. Survey Mon. 26, p. 81, i)l. 51, figs. l-(i, 1896. Bern,-, New Jersey Geol. Survey Bull. 3, p. 138, pi. 17, fig. 1, 1911; Torrey Bot. Club Bull., vol. 39, p. 395, 1912; vol. 44, p. 182, 1917. Liriodendron pinnalijidumf Knowlton mot Lesquereux), U. S. Cieol. Sur\-ey Twenty-first Ann. Rept. i>t. 7, p. 317, 1901. Liriodendron gnou-ii Knowlton (not Lesquereux'), idem. Leaves oblong, large, pinnatcly (iivided by narrow sinuses into two to four lateral lobes. Apc.x emarginate. Base truncate to somewhat cordate. Length along the midrib 7 to 9 centimeters and probably consiilerably more in some specimens, as one fragment measures 12 centimeters in wiilth. Width in perfect specimens about 9 centimeters. Lateral lobes ovate, with very acute tips, some narrowed proximad, giving them an almost obovate outline: intervening lateral sinuses narrow and deeply cut, some reaching nearly to the midrib, rounded. Some specimens have only two main lobes developed on each side and are then very similar to the typical modern leaf of Lirioden- dron tulipifera. In these specimens, however, the upper lobes are divided by a shallow sinus into two acute lobules. Other specimens show three lobes of equal magnitude on each side, and one of the best specimens from the Wood- bine sand has four nearly ec[ual lobes on each side, the basal and apical pairs being somewhat shorter than the medial pairs. This form of leaf is very suggestive of some species of Quercus, but its variations, as well as its vena- tion, show that it is related to Liriodendron. The petiole is preserved for a considerable length and is very stout, as is the midrib. There is one main secondary traversing each lobe and running directh' to its apical point. In addition there are one or more camptodroine secouflaries in each lobe which anastomose with branches from the main secondary, their number being dependent upon tlie relative wi- taceous and early Upper Cretaceous dicotyle- dons would result in transferring a number of forms to Trocliodendroides, wliich may .serve for the nH'('|)tiiin of any fossil s]iecies of tiio family. Truchudvndruidcs rhoniboidcus (Lesquereux Berry. I'late XXXVI, figure:.. Firuxf rhomboideu.1 Lesquereux. .\mi. .lour. Sci., 2ii ser., vol. 40, p. 96, 1868. Phyllites rhomhnidcus l.«8quereu.x, ( 'rotaceou.-' flora. \i. 1 12, ].l. 6, fig. 7, 1874. Knowlton, I'. S. (iool. Survey T\vciity-fir!1. This species was descrila'd l)y Li'siiikmcux from material foinid in (he Dakota sandstone at Decatur, Nebr. He at first referred it with a (|uery to FIcii.i, which it obviously does not represent. Siiliseciuently lie transferred it to " Berry, K. W., Am. Jour. Scl., -1111 s«r., vol. 80, p. «, !»». FLORA OF THIO WOOUBINF. SAXI) AT AHTHTRS BLl'FF, TKX. 167 Phi/llilis, aiul ill liis account of 1S74 ho sug- gested coniparisoMs with Smilas, Paliiirun, ami Populus. It has no s numerous, diverg- ing from the midrib at angles i>f .W to 40°, long ascending and eventually camptodrome, send- ing off small outwardly directed branches to tho miuginal te<-th. Texture coriaceous. This characteristics species with its coriaceous texture nmst have had i-atlu^r stiff, stri<'t loaves in life. It adtls to our Mora anottuT form of tho curious genus Ihwalqura, which is .so striking an elcMuont in the UjipiT Cretaceous and lower Eocene. Tho only known .Vmeriean species that resembles this form in any respect is Dtwdhjuia smithl Horry," of the Tuscaloosa and lilack Creek formations. D. smithi is much larger, with relatively broader leaflets, which have less prominently serrate margins and partly craspedodromo venation. Dewalquea insigniformis is, however, as its name indicates, very much like Dewalquea insir/nis Hosius and \'on der Marck," a promi- nent species in the Campanian and Maostricht- ian substages of Europe. D. Insidni.s has r«'latively broader, less prominently toothed leaflets, in some specimens as many as seven, and the venation is said to be craspedorlrome. The venation is, however, a character of slight value, for entire and toothed leaflets gener- ally occur together, and I imagine that De- walquea insignis is merely a serrate form of the associated Dewalquea haldemiana Saporta and Marion. If the latter had prominent ser- rate ttw?tli added it would be identical with Dewalquea insigniformis. Deivalquea insigniformis was described from a large number of fragmentary specimens ob- tained in the upper part of the Bingen sand in Pike County, Ark. .V characteristic leaflet is present in the early collections from the Wood- bine sand at Arthurs Bluff, Tex. Order ROSALES. Family CAESALPINIACEAE. Genus PALEOCASSIA Ettingshausen. Paleocassia laurinea Lesquereux. Plate XI., figures. Paleocassia laurinea I.e.s(|uereux, The flora of the Dakota group, p. 147, pi. 64, fig. 12, 1892. Berry, Torrey Bot. flub Bull., vol. ;J9, p. 396, 1912; U. S. Geol". Survey Prof. Paper 112, p. 100, pi. 23, fig. ], 1919. Leaflets ovate-lanceolate, subinequilateral, with a pointed ape.x and a cuneatebase. Length « Berry, E. W., Torreya, vol. 10, pp. 34-3.*, flg. 1, 1910. » Hosius, A., and Von der Marck, W., Palaoonlographica, vol. 20, p. 172, pi. 32, flgs. 111-113; pi. 33, fig. 109; pl..34, fig. 110: pl.35, fig. 123, 1S80. 168 SHOBTER COXrniBlTIOXS TO UKNERAL GEOLOGY, 1921. from 3 to 0 centimetci-s; maximum width, at or below the mitklio, about 2 rontimeters. Margins entire, somewhat irregular. Petiohih" short, curved, gradually enlarged proximad, about .5 milliiiietere in length. Midrib of medium size, curved. Secondaries thin, camp- tod rome. Tiiis species was described from specimens coUecteil in the Dakota sandstone of Kansas anti is present in the lower part of the Tusca- loosa formation in .Alabama. As inteipreted by Lesquereux. its describer. the remains rep- resented leaflets of a t'a.s'.v/a-like plant, although so far as I know all have been detachetl, the only basis for considering them leaflets rather than leaves being their slight iiic(|uii.it('ral form. A single entire leaflet is contained in the collection from Arlhui-s Bluff, Tex, It is identical with the type material from Kansas in size, outline, and venation, with the excep- tion that it is slightly wider (2 to n millimetei's), with a consequently somewhat fuller and more rounded base. Family PAPILIONACEAE, Genus COLUTEA Linn^. Colutea primordialis Heer. Colutea primorduilu Heer, Flora fossilis arctica, vol. 6, Abt. 2, p. 99, pi. 27, figs. 7-11 ; pi. 63, figs. 7, 8, 1882. I.e£quereux, U. S. Geol, Survey Mon. 17, p. 148, pi. 13, figs, 8, 9, 1891 11892], Newherr>', U, S, Geol, Purvey Mon. 20, p. 97, pi. 19, figs. 4, 5, 1896. Ilollic k, r. S, G«'<.1. Survey Mon. .=)0, [i. 84, pi. 32, figs. 14, 15. 1900. Berry, Torrcy Bot. \. 20, fig. 4, 191 1 ; Maryland Geol. Survey, Upper Crelaceous, p. 845, pi. 75, fig. 3, 1916; Torrey Br.l. ( hil. I'.ull , vol. 44, p. 184, 1917. This species was de.scril)cd from material found in the Atane beds of western Greenland and was subsequi-nlly rci-orded from the Dakota sandstone near iJclphos, Kans., the Raritan formation of New Jersey, and the Magothy formation of Marthas N'im'yard. Long Islan millimeters long; midrib stout, prominent; secondaries thin, five or six alternate pairs, branching from the midrib at tingles of about 50°, curving slightly upwai-d. anastomosing close to the entire margin. Lateral letiflets inequilateral, the outer limb of the lamina being slightly wider and fuller than the inner limb; petioles shorter than that of the terminal leaflet, 2 to 3 milhmeters in length, diverging from the latter at angles of about 70°; in outline and venation similar to the terminal leaflet, but smtiller and shownng a tendency to develop slight iiTeguhxrities in the margin, especially toward their tips. This species was named from its rather striking resemblance to the European early Tertiary species Rhus reddita Saporta,^' from Aix, in southeastern France. Severtd Creta- ceous species of lUuift have been described from specimens found in beds as old as the Wood- bine, the Dakota sandstone of Kansas having furnished three supposed species with pinnate leaves, one of which, Rhus uddeni Les(|uereux,'" was erroneously reported by Ward from the Cheyeiuie stmdstone at Belvidere, Ktuxs. The Cheyenne form proves to be a species of Saplndopsif<, and this mtiy ttlso be the botanic affinity of the type materitd of Khvs uddeni. A smaU-toothed species from the Cenoniiinian sandstone of Bohemia has been described by Velenovsky ^" :is Rhus cretacM, although tliis " Saporta, Gaston de, fitudes sur la vi5g^tatioii (lu sud-est de la France k I'epoque tertiaire, tome 1, p. 124, pi. 13, figs. 2, a, b, 1862. ^ Lesquereux, Leo, U. S. Geol. Survey Mon. 17, p. I.M, pi. -i", fiR. 2, 1892. » Velenovsky, Josef, Die Flora der IxJhini.scJie Kreideformation, pt. 4, p. 7, pi. 4, flgs.7-12, 1885. name was idready in use for a very difTerent Cretiueous species descril)ed by lleer^' from material obtained in the Senonian at Quedlin- burg, in Saxony, and recorded by Iliillick^'* from the Upper Cretaceous of Long Island. The Woodbine species is readily distinguishable from all the foregoing and adds a well-marked and probalily trifoliate Cretaceous form to this genus, wliich was so largely developed during Tertiary time. In the existing (lora Rhy.>< is a prominent element witli more tliaii l.'iO species, most of which are natives of warm temperate and tropical regions. A modern species \vith abnost identical foli- age is the South African Rhus villosa Linn6. Order RHAMNALES. Family RHAMNACEAE. Genus RHAMNUS Llnnfi. Rhamnus tenax Lesquereux. riate XL, figure 7. Rhamnu3 tenax Lesquereux, Am. Jour. Sci., 2d8er.. vol. 46. p. 101, 1868; The Cretaceous flora, p. 109. pi. 21, (ig. •!. 1874; The flora of the Dakotii group, p. 170, pi. 38, fig. 6. 1892. Engelhardt, Naturwiss. Geaell. I.sis in Dre.sden Al>h. 7, Jahrg. 1891. p. 101, 1892. Bartsch. Iowa Univ. Lab. Nat. Hist. Bull., vol. 3. p. 181. 1896. Berry, Torrey Bot. Clul) Bull., vol. 39. p. 398, 1912; U. S. Geol. Sur\ey Prof. Paper 112. p. 114. pi. 2.5. figs. 1, 2, 1919. Leaves ovate-lanceolate, acuminate, slightly etjuilateral. Length about 8 centimeters; maximimi width about 2 centimeters in the lower half of the leaf. The dimensions of these leaves are remarkably uniform in all the speci- mens from the Dakota sandstone of the West, as well as in those from the Tuscaloosa forma- tion in -Vlabama. Margins entire, curving in- ward somewhat abruptly to the petiole, which is stout, more or less curved, and about 1 centi- meter or slightly more in length. Midrib stout, curved, becoming thin in the acuminate tip. Secondaries numerous, thin, appro.ximately parallel, 12 to 14 subopposite to alternate pairs, branching from the midrib at angles of about 4.5° to 50°, curving slightly upward, campto- drome. This species, which was described many years ago by Lesquereux from material col- " Heer, Oswald, Ziir Kreideflora von Qiiedlinburg, p. 14, pi. 3, fig. 11. IS72. » HoUick, Arthur, U. S. Geol. Survey Mon. 50, p. S7, pi. 33, flg. 2,1907 170 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY, 1921. lect«d in the Dakota sandstone of southern Kansas and subsequently recorded by Bartsch from the same formation in Iowa, is repre- sented in the lower beds of the Tuscaloosa formation of western Ahibama and in the Woodbine sand at Arthurs Bluff, Tex., by leaves which are identical in all their characters with the t\'pe material and which seem to be closely allied to the Tertiary and modern forms of Rhammis. It has been reported by Engel- hardt from the Cenomanian of Niederschoena, Saxony, but this record may well be considered doubtful. Genus ZIZYPHTJS Adanson. Zizyphus lamarensis Berry. Plate XXXVI, figure 4. Zizyphus lamarensis Bern.', Torrey Hot. Cluh IhiU.. vol. 39, p. .398, pi. 31, fig. 1, 1912; U. S. (ieol. Survey Prof. Paper 112. p. 112, 1919. Leaves elliptical, 4.5 to 5 centimeters in length by .3 centimeters in maxinnini width, about midway between the apex and the l)ase, though slightly nearer the base ; base full and rounded; lateral margins full and rounded; apex rounded, slightly less full than the base; mar- gin with regular but shallow crenate teeth, becoming less prominent toward the base. Midrib slender but prominent, straight. Lat- eral primaries one on each side, diverging from the midrib at its extreme base at an acute angle (about 10°), thin, slightly curved inward above the middle, joining a secondary in the apical part of the leaf. Secondaries from the midrib two or three alternate thin pairs in the apical region, camptodrome; seconda- ries from the lateral [)rimaries five or six in number, on the outside, curved, camptodrome; the lowest secondary is longest and branches at the most acute angle (about 10°) and from the extreme base; each successively higiier secondary subtending a slightly larger angle and following a somewhat shorter course. Internal tertiaries more or less percurrent, marginal ones similar to the secondaries from the primaries in tlicir arrangement and course, thin and camptodrome. This handsome species of an undoubted Zizyphufi wius descrilx'd recently from very scant}' material obtained in the Woodbine sand of Texas. It is entirely distinct from any previouslv described Cri'laceous species and is much closer to some of the Tertiary and still existing forms. Zizyphus has not yet been discovered in the European Cretaceous, but it is represented in the Western Hemisphere by four or five well- marked types. The nearest to the present species is Z. groenlandiats Heer,'° which oc- curs in the Magothy formation on Marthas \'ineyard and in the Patoot beds of western Greeidand. It is of about the same size but relatively wider than the Texas form anil has a somewhat difl'erent venation and much coarser teeth. The species from the Magothy of New Jersey, Z. cUffwoodensui Berry,'"' is a larger, lanceolate, entire-margined form. The two species from the Oetaceous of Long Island, Z. eleyans ■" and Z. obloiu/iis,'' lislii'(l m I.S,S2, is as follows: ('. foliiH pulniulitt. profuDile trilohatia, lubu mu>li'> l'ius. pi. isii, Og. 13, IHIM. FLORA OF THE WOODBIXE SAND AT ARTITITRS BIATFF, TEX. 171 This description was based upon very frag- mentary material from the Atane beds of Green- land, from wliieli, nevertlieless, Ileer reeon- strueted the supposed outline of the perfect leaf. To judu:e by the specimens referred to this species by Lesquereux and Newberry it was an exceedin;ikolii ..'roup, p. 97, pi. 11, fig. 5, 1S02. Mnlapoennn fnlrifoHn ^T-esquereux) Knowllon, V. S. Geol. Suney bull. 152, p. 142, 1S98. Berry, Torrey Hot. Club Bull., vol. 33, p. 180, liioii; New Jersey (ieol. Sur\-ey Ann. Rept. for 190.'), p. 139, 1900; Torrey Hot. Club Bull., vol. 39, p. 399, 1912; V. S. Clcol. Survoy Prof. Paper 112, p. 122, pi. 21, 6g. 5, 1919. Leaves of relatively small size, lanceolate, falcate. Length about .") to 0 cciitimelors; maximum width, about halfway bclwiMMi the apex and the base, if anything slightly nearer the base, about 1.7 centimeters; from this point the blade luirrows to the lanceolate ba.se and gradually tapers to the extended acuminate tip. petiole not |)reserved. Midrib nnich curved, thin ilistad. Secondaries three or four pail's; tho lower suprabasilar and subopposite pair should possibly be termed lateral |)rinuiries, as Les(|uereux called tlieiii. Those lower secondaries are thin and l)raiich from tho mid- rib at angles of 1.")° or less, sweeping upward in a long curve, at length cainptodroino. Upper secondaries somewhat irregularly s|)aced, cainp- todroino. Tertiaries very fine and more or loss obsolete. Texture coriaceous but not thick. This attractive species is easily distinguished from other lauracoous forms, particularly from species of Cinnamomum, with which Lesque- reux originally compared it, by its slight in- (■(luilaterality and marked falcate form, as well as by the lack of deliniteness in the triple venation, tiie supral)asilar position of the so- called primaries, and tlie character of the tertiary venation. It was described originally from material found in the Dakota sandstone near l)<"lp]ios, Kans., and subso(piently was discovered by me in the Msigoth}- formation of Xew Jersey. Some of the specimens from the lower part of tho Tuscaloosa formation of Ala- bama, though the material is not extensive, are comi)leto and are entirely characteristic, as is the single specimen discovered at Arthurs Bluff, Tex. It may readily be distinguished from Malapni una fiorrellen.si9 Berry ,-''^ of the Upper Cretaceous IJlack Creek, Eutaw, and Ripley formations, by its suprabasilar primaries and cuneate base. Genus OREODAPHNE Nees. Ureodaphnc alabamcnsis Berry. Plate XXXVII, figure 1. Ormclaphne alabainensis Berry, Torrey Uot. Club Bull., vol. 39, ]). •100, pi. 32, 1912; U. S. Geol. Survey Prof. Pui>er 112, p. 119, \>\. 19, figs. 3-5, 1919. Leaves of large size, ovate, from 13 (o 20 cubic centimeters in length and from 4.75 to 7 cubic cciitiincloi-s in mii.xiimmi width, which is at a point midway between t!ie apex and the base. From tiie |)oint of greatest width the miirgiiis cur\c, both distad and proximad, in a V(>i\ full cni\ e. inurowing rather abruptly to the assil species of this genus and from Ciiuiainomum. Cocculus, or other genera having somewlial similar lea%es, with whicli it might be compared. Texture coriaceous. This fine large species is represented at Arthurs BhifT, Tex., by fragmentary but char- acteristic specimens. The description was largely drawn up from abundant and complete material from the Tuscaloosa formation of western .Vlabama. The spi'cimens show con- siderable variation in size and some in outline, the leaf being widest either nearer to or farther from the base. In the latter form the distal part is more fully i-ounded and abruptly con- tracted to the acuminate tip, while the base is more gradually narrowed and finally cuneate rather than rlecurrent. In the former the apical part is more gradually narrowed and the base is full and rounded abruptly, decurring to the petiole. This species is markedly different from other described fossil forms but may be matclied by several modern tropical American species of Oreodaphnf. The genus Oreodaphne of Nees, which is exclusively American in the existing flora, is made a subgenus of Ocotea Aublet by Pax in Engler and Prantl's '' Die natiirlichen Pfl^mzenfamilien." The genusOcoteu. which for paleobotanic purposes may be considered as composite, has about two hundred modtrn species occurring chiefly in the American Tropics and ranging from southern Florida to Brazil and Peru hut having some representa- tives (subgenus MexpUodaphne Nees) in the Canary Islands, South Africa, Madagascar, ami the Mascarene Islands. The single existing American species reach- ing the I'nited States whose habit and environ- ment may be taken as typical for the whole genus is found in Florida, southward from Capes Canaveral and Romano, along the shores and islands, except on some of the western keys, making its best growth in the rich, moist ham- mock lands near the coast. Oenus CINNAMOMTTM Blume Cinnamomum nowberryi Herrr. Pluti' XX.XIX, liKur.-;!. Crnrumwnium sezunnciisi- Ilcer. Flora fotsilin arctira. vol. ti, Al.t. 2. p. 77. pi. lit, liR. «; pi. :!:$. figs. 11. 12, 1882 (not. Wiilcleti; vol. 7. p. :«l, pi. 11, fig. la. 1883. LeBqucreux, The llora of the I»akota group, p. 107, pi. 12. lig. 7, 1892 (not lig. (i). l)a\vson. Roy. Soc. Canada TraiiK., Istser., vol. 2. sec. I. p. 6-1. pi. 13, fig. .")S, 1891. Hollick. Torrcy Bot. Cluh Bull., vol. 21, p. h'i. pi. ISO, figs. 5, 7, 1891. Penhallow. Roy. Sor. Canada Trans.. 2d ser.. vol. 8, sec. 4. p. -It). 1902. Hollick. New York State Mu.s. Kilty-IUlli .\iui Kept., for 1901. p. r.50, 1903. Cinnamomum miermcdium Newlierry. .Smitli, On the ge- ology of the Coastal Plain of Alabama, p. 348, 1894 (nomen nudum) (not Ettingshaiisen). Newberry, The fiora of the Amboy clay.s. p. 89. pi. 29. figs. 1-i. 1896. Berry. New.l ersey Gcol. Sur\-ey .\nn. Rept. for 190.'). p. 139. pi. 20, figs. 2-6. 1906; Torrey Bot. Club Hull., vol. 33. p, 179. pi. 7. figs. 3. 4; vol. 37, p. 27. 1910. Hollick, The Cretaceous flora of southern New York and New England, p, 74, pi. 29. fig. 7; pi. 30, figs. 1, 2, 1906. Cinnamomum neubcrryi Berry. Torrey Bot. Cluli Bull., vol. 38. p. 423. 1911; New Jersey Geol. Sun-ey Bull. 3. p. 150. pi. 16, fig. 3, 1911; U. S. Geol. Sur- vey Prof. Paper 84. pp. b4. 117. pi. 9. figs. 12, 13; pi. 21, figs. 9-11, 1914; Maryland Geol. Survey. Upper Cretaceous, p. 860. pi. 71, fig. 6. 1916; U. S. Geol, Survey Prof. Paper 112. p. 118. pi. 21, figs. 6-9, 1919. Cinnamomum n. sp.? Knowlton, U. S. (ic nora flcr biihiniflciio Krddo/ormatlon, I't. 2, P-in.r , . , asa. think that IloUick was entirely justified in removing this form from Paliuru--<, but I am not sure that it is a Ciiniamomum. altliough it appears to be a lauraceous foyn. If a Ciinia- momum, as is perfectly possil)le, it should probably be regarded as a variant of the con- temporaneous Cinnamomum newherryi, from which it differs merely in its irregularity of outline. Genus LAITRUS of authors. Laurus plutonia Heer. Plate XXXVIII, figure 5. Laurus plutonia Ilocr, Flora foasili.s arctica. vol. 6, Abt. 2, p. 75, pi. IS), figs. Id. 2-1. pi. 20, figs. 3a, 4, 5; pi. 24, fig. 6b; pi. 28, figs. 10, 11; pi. 42, fig. 4b, 1882; vol. 7. p. ,30, pi. 58, fig. 2; pi. 02, fig. la, 1883. ?Velenovsky, Die Flora der bolimische Kreideforma- tion, pt. 3, p. 1, pi. 4, figs. 2-1, 1884. Lesquereux. U. S. G«ol. Survey Mon. 17. p. 91, pi. 13, figs. 5. 0; pi. 22, fig. 5, 1892: Minnesota Cieol. and Nat. Hist. Sur\'ey, vol. 3, pt. 1. p. 14, jil. A, fig. G: pi. B, fig. 5, 1895. Newberry. U. S. Cieol. Survey Mon. 2G. p. 85. pi. 16, figs. 10, U, 189G. ?Fri6 and Bayer. Naturw. I.andos. Bohmen .Vrchiv, vol. U, No. 2, p. 130, fig. 94, 1901. Berry, New York Bot. Garden Bull., vol. 3, p. 79, pi. 1, figs. 9-11, 1903: Torrey Bot. Club Bull., vol. 31, p. 77, pi. 3, fig. 1, 1904: vol. 33. p. 178, 1900; vol. 39, p. 401. 1912: New Jersey Geol. Survey Ann. Rept. for 1905. pp. 138, 139, UIOG: U. S. Ck-oI. Sur- vey Prof. Paper 84. p. 52. pi. 11. fig. 2; pi. 12. fig. 6, 1914: Maryland Geol. Sur\'ey. Upper Cretaceous, p. 861, pi. 71. fig. 5, 191G; U. S. Geol. Survey Prof. Paper 112, p. 123, 1919. Hollick. U. S. Geol. Survey Mon. 50. p. SO. pi. 27, figs. 9. 10, 1900: New York Bot. Garden Bull., vol. 8, ].. 102, pi. 1(;9. figs. 3l,5. 1912. Leaves lanceolate, usually tapering almost equally in Ixilh directions but some specimens less acute at the base. Length, 7 to 11 ccnti- m(>ters: greatest width, 1.5 to 2.5 centimeters. Midrib fairly stout. Petiole short and stout, 6 to 15 milhmeters in length. Secondaries slender, eight or more alternate pairs, camp- todromo. This species wa.s described by lleer from material eolieeled in the .Vtane Ix'ds of western Greenland, and a large numl>er of somewhat variable and fragmentary specimens were figured. Snbse(|nently it wiis recorded from a very largo number of Cretaceous plnnl beds, 80 that its range both geographic and geologic is rather exUWisive. \ number of these records are not entirely above sus|)i(ion, and this FLORA OF THE WOOPBIXF, SAXP AT ARTHl'RS BLrKF. TKX. 175 appptirs to he cspecinlly true of the forms from the Conomnniiin of Bohemia iih-ntilied hv \'elenovsky. fMiinm philonia is uncommon in the Karitan formation, and I have found it only near tlie top. It is abundant in tlic ovorlyinjl Maijothy formation from New Jei-sey to Maryhmd. In the southern Coastal Plahi it occurs in the Middendorf arkose member of the Black Creek formation of South Carolina and ranges from the base of the Tuscaloosa formation upward into the Eutaw formation in the Alabama area. It was identified by Ward from the Cheyenne sandstone of Chatman Creek, Kans., but his material, which I have studied, is not this species but represents leaflets of Sapindopsis. A single complete and characteristic l(>af and several fragments are contained in the collections from Arthurs Bluff, Tex. Lauras antecedens Lesquereux? Lauras antecedens Lesciuereiix, U. S. Geol. Survey Mon. 17, p. 92, pi. 11, fig. 3, 1891 [1892]. HoUick, U. S. Geol. Survey Mon. 50. p. 80, pi. 28, figs. 9, 10, 1900. This species, the type localitj^ of whicli is simply '■■ Dakota sandstone of Kansas (Lacoe collection)," is at best of doubtful validity. It was described by Lesquereux as follows; Leaf membraneous, lanceolate, jrradually tajjering to the apex, narrowed to the l)ase, not dccurri'iit, soinewliat curved to one side, entire, irregularly undulate; median nerve thick; secondaries oblique, curved, parallel, hut of unequal thickness and distance, caraptodrome. The leaf is n centimeters long, 2.5 centimeters broad below the middle, slightly inequilateral by the partial contrac- tion of the l)orders on one side, and is not gradually nar- rowed to the petiole but somewhat rounded in narrowing to it. Its precise relation is not satisfactorily ascertained. The Texas material is fragmentary and of doubtful identity. Genus LATJROPHYLLTJM Goeppert. Laurophjilum minus Newberry. Laurophyllum minus Newberry, U. S. Geol. Survey Mon. 26, p. 87, pi. 17, figs. 7-9, 1895 [189G]. Berry, N. J. Geol. Survey Bull, '.i, p. 149, 1911: Torrey Bot. Club Bull., vol. 39, p. 102, 1912. This species, which is of doubtful validity, was described from material collected in the Raritan formation of New Jersey, and I have found it only in the upper part of that forma- tion. In the absence of venation characters in both the type and later collected material its identification is always more or less uncer- tain; and it may represent a variety of Imuhis plittonia Ileer or some of the forms that have been referred to Miirica loiifja Ileer, although in general it is wider than the latter and more elongated and less symmetrical than the for- mer. A single specimen is present in the collec- tion from Arthurs Bluff, Tex. Order MYRTALES. Family MYRTACEAE. Genus MYRTONITJM Ettingshausen. Myrtonium geinitzi (Heer) Berry. Myrtophyllum (jeinUzi Ileer, Kreidoflora von Moletein in Mahren, p. 22, pi. 11, figs. 3, 4, 1872; Flora fosailis aniica, vol. 3, Abt. 2, p. IIG, pi. 32, figs. 14-17, 1874. Krif, Naturw. Landes. Bohmen Archiv, vol. 4, No. 1, l)p. 18, 94, 1878. Hollick, New York Acad. Sci. Trans., vol. 12, p. 236, pi. 6, fig. 2, 1893. MijTtophjUum wardcri Lesquereux, U. S. Geol. Sur\ey Mon. 17, p. 136, pi. .53, fig. 10, 1892. Uollick, U. S. Geol. Survey Mon. 50, p. 97, pi. .35, fig. 13, 1906. Eucalyplusf angusli/olia Newlierry, U. S. Geol. Survey Mon. 26, p. Ill, pi. 32, figs. 1, 6, 7, 1896. Hollick, New York Bot. Garden Bull., vol. 3, p. 408, pi. 70, figs. 8, 9, 1894; U. S. Geol. Survey Mon. .50, p. 95, pi. 35, figs. 9, 14, 15, 190G. Eucalyptus geinitzi Ileer, Flora fosailis arctica, vol. G, Abt. 2, p. 93, pi. 19, fig. Ic; pi. 45, figs. 4-9; pi. 46, (iga. 12c, d, 13, 1882. Engelhardt, Naturwiss. Gesell. Isis in Dresden .\bh. 7, p. 102, 1891. Lesquereux, U. S. Geol. Survey Mon. 17, p. 138, pi. 37, fig. 20, 1892. Newberry, U. S. Geol. Survey Mon. 20, p. 110, pi. 32, figs. 2, 12, 15, 10, 1890. Kra-sser, Beitriige zur Kenntniss der fossilen Kreide- flora von Kunstadt in Mahren, i). 22, 1896. I'ric and Bayer, Naturw. Landes. Bohmen .Xrchiv, vol. 11, No. 2, I . 142, fig. 110, 1901. Berry, New York Bot. Garden Bull., vol. 3, p. 87, pi. pi. fig. 3, 1903; Torrey Bot. Club Bull., vol. 31, p. 78, pi. 4, fig. 5, 1904; vol. 33, p. 180, 1906; vol. 34, p. 201, pi. 15, fig. 4, 1907; vol. 37, p. 26, 1910; vol. ;», p. 402, 1912; New Jersey Geol. Survey Ann. Kept, for 1905, p. 138, 1906. Hollick, U. S. Geol. Survey Mon. 50, p. 96, i)l. 35, figs. 1-8, 10-12, 1906. Hollick, New York Bot. Garden Bull., vol. 8. p. 166, pi. 180, figs. 1, 2, 1912. Berry, U. S. Geol. Siu^ey Prof. Paper 81, p. 56. pi. 13, figs. 8-12; pi. 14, fig. 1, 1914; New Jereey Geol. Siu^ey Bull. 3, p. 189, 1911; Maryland Geol. Sur- vey, Upper Cretaceous, p. 870, pi. 81. figs. 1-5, 1916; r. S. Geol. Sim'ey Prof. Paper 112, p. 126, pi. 28, fig. 8, 1919. 176 SHORTER CONTRTBrTlONS TO GENERAL GEOLOGY, 1921. This widespread and characteristic Upper Cretaceous species is fi)iiiid in tiiis countrv from the base of the Karitan formation of New Jersey upward into the Black Creek formation of tlie Carolinas. Abroad it is common in the CVnomanian and ranges upward into the Turonian. The type locality is in the Ceno- manian of Moravia. l>ut the species has also been reconlcd from tlic Atane Ix'ds of western Greenland and from the Dakota sandstone of the West. In the .Vlabania Cretaceous it has been collected only from the lower |)art of the Tuscaloosa formation. Several characteristic specimens are present in the collections from Artiiurs BlufT, Te.\. This species was recorded l)y Ward from tlie Cheyenne sandstone at Chatman Creek, Ivans., but the material upon which tlie record was based is referable to the genus Sapindopsis. Order UMBELLALES Family ARALIACEAE. Genus ABALIA Llnn€. Aralia wellingtoniana Lesquereux. Plate XXXVII, figure 3; Plate XXXVI II, figure.s 3. 4. Aralia wellingtoniana Lesquereux, U. S. Geol. Siu-v-ey Mon. 17, p. 131. pi. 21, fig, I (pi. 22. figs. 2. 3. is not this species but Aralia saportana Lesquereux), 1891 [1892|. Newl)err>-, U. S. Geol. Survey Mon. 35, p, 114, pi. 26. fig. 1. 189.1 [1896], Berr>'. New Jersey Geol. Sur^'ey Bull. 3, p. 202. pi. 2h, fig. 7. 1911; Torrey Hot. Club Bull., vol. 39. p. 402. 1912. Aralia concinna Newberry, U. S. Geol. Survey Mon. 16 IC. p. 114, 1895[1896|. . Aralia wellingloni/ina laughanii Knowlton, U. S. Geol. Sur\-ey Twenty-first Ann. Rept..pt. 7, p. 317, 1901. This handsome species is descril)ed l)y I>es(|uereux as l)eing palmately three to live lobed, but it certainly seems significant that all the forms from the Raritan formation are three-lolx'd and that the (ivc-lol)ed forms from the Dakota sandstone referred by Les<|uereu.\ to this species are indistinguishable from his .species Aralin Mnporlonn, which occurs at the same horizon and. in part at least, at the same locality. 'lliis is the most abimdant form collected at Arthurs BlufT, Te.\.. there being 15 specimens in the one small collection made by Stanton and St^'phenson and as miuiy more in the old collections made by Hill and \'aughan. These are all trilobate, and the majority liave toothed nuu'gins and agree exactly witii tiu' Karitan leaves of this species and with tiu> trilobate leaves frcuu the Dakota sandstone like the one figured l)y Les(|uereux on his Plate XXI, figure 1. In the light of our present knowledge Aralia welliinitoiild may be redescribed in tlie following terms : Ijcaves variable in size, 10 to 20 centimetei"s in length by S to lo centimeters in maximum width from tip to ti|) of tiie lateral lobes; average size about 15 centimeters in length by 11 centimeters in wiiltli: coriaceous, palmately deeply trilobate, with a rapidly narrowed and more or less extended decurrent base; lobes long, lanceolate, widest in the middle and narrowing below, somewhat abruptly acumi- nate, the medium slightly the longest, diverg- ing at an angle of about 30°, sepiraled by sinuses extending more than halfway to the base, narrowly rounded: margins entire below and for varying distances upward, sometimes throughout, generally passing gradually into dentate-serrate teeth, one to each secondary or less, prominent in some sj)ecimens, where they arc mor(> or less extended and directed upward, separated by wide, shallow sinuses. Primaries stout, suprabasilar, the median slightly larger than the laterals. Secondaries numerous, thin, regular, subparallel. ascending, as the angle of their divergence from the primaries averages about Xl^. luit slightly curved in tli(>ir course, ultimately craspedo- drome in the distal parts of tlie leaf, \vlii>re the margin is idoilicd. and < ani|)t(idr()nie in the basal half of tlie leaf, wiiere (lie margin is entire. .Vreolatioii indistinct, reticulate, of r|uadragonal or polygonal meshes. The smaller leaves are relatively shorter and broader, with less extended lobes and more open and less deep sinuses. The present species was confused by Ward'" with what was subsequently dilferciil iated as Aniliii coltoiKldli iisis Berry,'" of the Tus- caloosa formation, which has shorter, more conical lobes, a broadly roundi'd base, and more crenate marginal teelh. •■ Wuril, L. F., In Smith, E. A., neology of Uic C, I 3, IDIU. FLORA OF THK WOdimiNK SAND AT AHTIUHS BLUFF, TEX. r Tlio present species is also very similar to Aralia (hcurnns Velenovsky," from the Cenomanian n' floras^, p. (il. pi. 8, figs. 1. 2: pi. 9, figs. 1, 2, 1883. " Aralia trellingtoniana Lesquereux, Flora of the Dakota group, p. 131 (part I. pi. 22, figs. 2, 3 (not pi. 21, fig. 1). 1892. This species was described by Lcsfjuereu.x as follows : Leaves large, subcoriaceous. triple-nerved and five- lobate by di\Tsion of the lateral norvos, fan-shaped in out. line, narrowed in a curve or broadly cunoate. and decur- ring to a long, slender petiole; lobes narrowly lanceolate or linear-lanceolate, acute or blunt at the apex, equally diverging, distantly dentate from below the middle up- ward; secondary nerves subcamptodrome. This beautiful species is known by numerous finely pre8er\-ed specimens. The leaves. 9 to 20 centimeters long from the top of the petiole to the suinit of the middle lobe, are of the same width botwoen the points of the lower lateral lobes; the petiole is long and comparatively slender, though appearing thick upon one of the speci- mens, probably enlarged and flattened by compression. The preserved broken part on one of the leaves measures 5 centimeters. The lobes cut down to about two-thirds of the leaves are narrowly lanceolate, slightly narrower near the obtuse sinuses, equally diverging, the lower lateral ones much shorter, curved down, and decurring to the base of the leaves. The leaves, triple-nerved from the di\Tsion of the primary nerves a little aliove tlie Itase, become tive-ner\ed from the forking of the lateral nerves at a short distance from their base. The secondary veins emerge at an acute angle of 30°, curve in ascending to the borders, and sometimes enter the teeth by their ends; the upper more generally follows close to the borders in fes- " Velenov.>iky, Joset, Die Flora dcr bohniische Kreidcforniation, pt.3, p. 11,ijI.4, fig5. 5-7, 1S84. « Knowlton, F. H., U. S. Geol. Suney Twenty-first Ann. Kept., pt.7, 317, 1901. toons, emitting under the teeth short branches which enter them. There arc not an>- Intel nieiliatc tertiary veins, but the nervilles are strong, often continuotis. ana,-jiiig in the middle of the areas and forming by siilidivisinns a small quadrangular aroolalion. The material from Arthurs Bluff, Te.x., is scanty and nut |)ositively determined. Family CORNACEAE Genus CORNOPHYLLUM Newberrj- Cornophyllum vetustum Newberry. Cornophylbim iiliislinn Ni-whorry The flora of the Amliov cla\'s. p. nil. pi. 19, n.;. 10, 1890. Berry, Xew .lersey Geol. Survey Bull. :!, ]i. l9(i. 191 1 ; Torrey Bot. riul> Bull., vol. 39. p. 404, 1912; U. S. Geol. Survey Prof. Paper 112. p. 129. 1919. Leaves elliptical, 7 to 8 centimeters in length by about 4 centimeters in maximum width, with an acute apex and base, the base slightly decurrent and ine<|uilateral. Margin entire, very slightly and inconspicuously imdulate. Midrib slender and straight. Secondaries slender, about seven pairs, opposite or alter- nate, branching from the midrib at angles of about 45° and strongly curved upward, ap- pro.ximatcly parallel and camptodrome: they increase in length from the ape.x to the base, the lower ones sweeping upward in stnmg arches parallel with the margin and all drawn inward toward the apex. With the exception of the delicate and .some- what fle.xuous character of the venation, these leaves are strictly comparable with those of Cornus, good species of which, very similar to this species, occur in the Dakota sandstone of the West, in Greenland, and in the Magothy formation of Maryland. This species is found also in the Karitan formation of Nf^w Jersey and tiie Tuscaloosa formation of Alabama. Order ERICALES? FamUy ERICACEAE? Genus ANDROMEDA Liiine .\ndromeda novaccacsareae Hollick. Plate XXX\^II, figure 1. Andromeda noi21. Lea res small, thick, and entire, \vi(!i stout petioles and inidrihs and <)l)scure secondary venation whicii is inunersed in tlie thick lam- ina. Length 2.5 to .5 centimeters; width 0.9 to 1.3 centimeters. Venation, where visible, showing numerous parallel, camptodrome, relatively long and thin secondaries, which branch from the midrib at acute angles. Though the majority of these leaves are equally acuminate at both ends, there is considerable variation in this respect, and a well-marked tenilency is siu>^^^l in many specimens tliat are relatively broader, especially in the u|)pcr lialf, toward an obtusely rounded apex, the terniina- tion of the midril) showing as a small nuK ronate point. The base in these forms grachiaiiy nar- rows to the stout petiole. The variations in outline of this species are well shown in the figures reproduced in Newberry's monograph. Typical leaves of this s])ecies are not un- common at Arthurs Bluff. Tex. It was de- scribed originally from specimens found in the upper part of the Raritan formation of New Jersey and has subse(|uently b(>en recognized in the Magothy, Black Creek, and Tuscaloosa formations of the southeastern Atlantic Coastal Plain and in th(^ Bingen sand of Arkansas. Somewhat similar obovate leaves of very coriaceous texture, formerly confused with this species, are soraewliat younger and have recently been transferred by me to the genus Etipltorhioiihi/lhim. It may well be questioned whether this and the following species of Andromeda should be referred to the Ericales. Certainly the |)resent form has numerous points of contact with the leaves of Eiujenin, which has been positively recognized in the lower Eocene (loras of this general region. Andromeda snowii Lesquereux. An/lronwla snovni TiCsquereux, U. S. (ipol. Survoy ^ton. 17, p. 117, pi. 17, (ic. 1'!, 18!H (IWI'-'l- Herry, Torrcy Hot. t'lub Hull., vol. :VI, ]>. inr>, 1!)12. Leaves small, entire, coriaceous, lanceolate, broadest in the middle and ec|ually acute at both ends. Length 4. .5 centimeters; maximum width 1.0 centimetci-s. Midrib stout. Second- aries obli(|ue, regularly spaced, suhparallcl, camptodrome. This species, basal portion of the Eutaw formation and into the Coffee sand member of that formation in Tciuiessee. It is undoubtedly present, hut not common, in the collections from Arthurs Blufl', Tex. Order RTJBIALES. Family CAPRIFOLIACEAE. Genus VIBURNUM Linne. Viburnum robustum Lesquereux. Plate XXXIX, figure 4. Viburnum robushtm Lesquereux, U. S. Geol. Survey Mon. 17, p. 120, pi. 20, figs. 4-6, 1891 [1892]. Knowlton, V. S. Geol. Survey Twenty-first Ann. Kept, pt. 7, p. 317, 1901. Herry, Torrey Bot. Club Bull., vol. 39, p. 40.5, 1912. Leaves shortly and broadly ovate, generally widest below the middle, narrowing upward to the obtuse tip. Base cuneate, slightly decur- rent to the thick petiole. Margins entire, often somewhat undulate. Texture coriaceous. Length 7 to 10 centimeters: maximum width 5 to f) centimeters. Petiole long and stout, about 3 centimeters in length. Midrib stout. Secondaries stout, ascending, somewliat irregu- larly spaced, rather straight, camptodrome or hrachydrome. This species is represented by characteristic specimens from .Vrthurs Blufl', Tex. It was described originally from material collected in Ellsworth County, Ivans., and is not known from other areas. POSITION UNCERTAIN. Genus TRICALYCITES Newbcrrj'. Tricalycitcs pupyruceus HoUick. Plato XT., figure 9. Triralyriles pa pyraccu-ii lloUick, Torrey Hoi. ('lul> Hull., vol. 21, p. 63, pi. 180, fig. 8, 1894. Newberry, U. S. Geol. Survey Mon. 26, p. 132, pi. 40, figs. 30-38, 1896. llollick. New York Acad. Sci. Annals, vol. 11, p. 423, pi. 37, figs. 1, 2, 1898; Now York Bot. Garden Bull., vol. 2, p. 405, pi. 41, fig. 3, 1902; U. S. Geol. Survey Mon. 50, p. 109, pi. ,5, figs. 8-12, 1906. Berry, Torrey Bot. Club Bull., vol. 31, p. 81, pi. 1, fig. 4, 1901; vol. 39, p. 40r,, 1912; New .Tersey (ieol. Survey Ann. Kept, for 1900, p. 139, 1900; New Jersey Geol. Survey IJull. 3, p. 221, 1911; V. S. Geol. Survey Prof. Paper 112, p. 137, pi. 28, figs. 1-'), 1919. This very characteristic tri-alate fossil is abundant iii the middle and upper parts of the Raritan formation of New Jersey. It occurs sparingly in tlio overlying Magothy formation and is very common in the lower part of the Tuscaloosa formation in western Alabama. It is abundant at .\rthurs BlufT, Tex., the present cullectiiin containing eight typical specimens, some of them complete. They are in exact agreement with the Tuscaloosa forms and demonstrate what is discussed at length in my account of tite Tuscaloosa flora, that the approximately parallel longitudinal venation of the wings is I'eaily a more or less forked and anastomosing venation, thus allying these fossils in a remote way with such modern genera as Vatica, of the Dipterocarpaceae. This form, despite its uncertain botanic affinity, is an important stratigraphic type, readily and surely recognized at all times. It characterizes the Tuscaloosa, Woodbine, Rari- tan, and Magotliv formations but has never been discovereil in the Dakota sandstone. Genus CARPOLITHUS of authors. Carpolitlius sp. 1. A coriaceous, ovate, concavo-convex scale or fruit f)f unknown botanic aflinity, represented by a single specimen in the Woodbine sand at Arthurs BlufT, Tex. It is of no \ alue, either geologic or botanic, but is entirely unlike previou.sly described forms. 180 SHORTER CONTRIBUTIONS TO GENERAL GEOLOfiY, 1021. Carpolithus sp. 2. Plate XL, figure 1 . A coriaceiius vnho of a capsule or pod, latiTally roinpivssed, oval, and acumiiiato at btitli onds. About 1.5 contimetcrs in longtli !ind 8 millimeters in ma.ximmn width, in the median region. Represented by a single speci- men in the Woodbine sand at .Vrthui-s Blull, Tex., and of unknown botanic aflinity. Sug- gestive of a valve of some coriaceous, single- seeded leguminous form. (arpolithus sp. 3. Plate XL, figures 2-4. A rhomboidal or obovate object with a some- what transversely wrinkled, broadly rounded apex and a cuneate base, marked below the transvereely striated apical arc by ascending, closely spaced, acutely forking subparallel veins. Of unknown botanic aflinity, probably bracteate in character. Represented by three specimens in tlic Woodbine sand at Arthurs Bluff, Tc.K. Figure H — Cut in Ibe Woodbine sand near .\rihurs lllull. Tex. PLATES XXXVI— XL. 181 V. S. GKDLOOICAI, siKx ^^ I'HOFESSION'AI, I'AI'KU 1:>1) Pl.ArE X.XXVI P/ V \ ^r). ~~7 %^ - i 4. : / ■ V ,.\, I, Brarhyphyll qtien^iroliiin l.esqiii;r(Mjx ritssiL iM.Axrs fiiom tim: wikidiumc sand. urn macnHiirpuiii funiiosum Herry; 2. P.Klozaniil.-s lanrrohiins (I.hidliv and Hiiiton) V. Ilrauii INewberry; 4. ^izypJius lamareiisia Berry; 5, Trochodcmlroid.^s rhomboid<;iis llerry; 6, S 3, JJri4)(J4>iidroii i^rciilia liigiibrLs U. s. OKOi.(ic;i('Ai. ,siii;vi:\ I'llOFKSSIONAT. r.VPRR 13!) n.Al'K WWII 1 H. ■'-/^> N I i KOSSII. I'l.ANTS KMOM TIIK WOODIIINK S\N|) .1 ,. H..,r. ■' III,,,. ,, ,l,liiir,„i„;. Il.iiv; 1. Arriliii wi-lliiii!l"iii I I .i-M|inTi>in. V. S. «EOI.III.i< Ai >t lis t:\ PUOrKSSIONAI. PAPEU 129 IM.AI K XXXVIII >/ (.. "^^ y 4 KOSSIL I'LANIS KltOM 'mi-: \\(MH)11IiM: SAiM). I, Amlromcfla iiovui^caesarriir I!ollisiiiirrcM\j KnoultDii; .'J. 1. Ar.-ili;i \vi*Ilint't^- \ KOSSII, l>I,\NTS IMl)\1 Till; WOODIIIM-; SAM). I, Fii-U« •IliphlKiKxnoiilm (IIcit) Hnrry; 2. Di.isriynii liriiiirinvii lliwr; .1, Ciiiiiutiioiiliim li.» li(rr> i lli-rr); 1. \ ilMiriiiiiii rulllHllJlll l.4!K<|lltTrMi; 5. Myril.'u InlIKU lllrrr) llftT. r. s. OFoinr.irAi. SURVEY PnOFESSIOKAI. TArKK 129 I'l ATK Xl. ^- f 1* i^ / ^^ FOSSII, PLANTS KHOM TIIK WOODlilNI'; SAM). 1, Carpolilliua sp. 2; 2. :!, 1. ('.:ir|iiililliiis si>. 3; 3. Cissilcs fi)rniosiis llc-i-r; 6. M;if:iiolia sppciiisu H