yE926 .B452 Berry, Edward Wilber The physical conditions and age indicated by the flora of the alum bluff for- mation V DEPARTMENT OF THE INTERIOR Franklin K. Lane, Secretary UNITED STATES GEOLOGICAL SURVEY George Otis Smith, Director Professional Paper 98— E THE PHYSICAL CONDITIONS AND AGE INDICATED BY THE FLORA OF THE ALUM BLUFF FORMATION BY EDWARD WILBER BERRY Published May 27, 1916 Shorter contributions to general geology, 1916 > (Pagea 41-59) WASIIIXGTON GOVERNMEXT PRINTING OFFICE 1916 DEPARTMENT OF THE INTERIOR Franklin K. Lane, Secretary UNITED States Geological survey George Otis Smith, Director Professional Paper 98 — E THE PHYSICAL CONDITIONS AND AGE INDICATED BY THE FLORA OF THE ALUM BLUFF FORMATION BY EDWARD WILBER BERRY PubUshed May 27, 1916 Shorter contributions to general geology, 1916 ( Pages 4 1-59 ) LIBRARY N'K\N' ^■o«K BOTANICAL (jAliDtiN WASHINGTON GOVERNMENT PRINTING OFFICE I'jie CONTENTS. Page. Introduction 41 Geology of the deposits 41 Character of the flora 43 Age of the flora 44 Systematic account of the flora 46 ILLUSTRATIONS. Page. Plate VII. A, Plant-bearing beds in the Hattiesburg clay near Raglan, Forrest County, Miss.; B, Plant-bearing beds at the tj^pe locality of the Alum Bluff formation, Alum Bluff, Apalachicola River, Liberty County, Fla 56 VIII-X. Fossil plants from the Alum Bluff formation at Alum Bluff, Fla 57-59 Figure 8. Section of Ilattiesburg clay near Raglan, Miss 42 II THE PHYSICAL CONDITIONS AND AGE INDICATED BY OF THE ALUM BLUFF FOMIATION. THE FLORA By Edwakd WiLBKR Berry. INTRODUCTION. Tlie present paper has for its purpose the description of a small ilora collected from the Alum Bluff formation, representing a horizon hitherto unrepresented paleobotanically in southeastern North America, and the discus- sion of the beai'ing of this flora on the physical conditions of deposition and the probable age of the deposits. GEOLOGY OF THE DEPOSITS. The ^Uum Bluff formation was named from the bluff of that name on the east bank of the Apalachicola River, about 25 miles l)elo\v Chat- tahoochee or River Junction, in Liberty County, Fla.i (See PI. vii, B, p. 56.) It is, according to present knowledge, the uppermost formation of the Apalachicola group. It comprises three membci's, which, named in ascending order, are the Chipola marl member, the Oak Grove sand member, and the Shoal River marl member. The Chipola marl, which is a thin yellowish clay marl at the base of the formation as defined by Matson and Clapp, caiTies a very extensive and well-preserved marine fauna. It was named from Chipola River, in Calhoun County, Fla. The Oak Grove sand, stratigraphicall}' mtermediate between the Chipola and Shoal River, is a thin, highly fossilifcrous gray or greenish fine sand named from Oak Grove, on Yellow River, and not represented by a litho- logic unit at Alum Bluff. The Slioal River marl, the highest known fossiliferous member ' The geology of this region is lully discussed in the following publica- tions: Matson, G. C, and Clapp, F. G., A prelimiuair report on the (jeology of Florida: Florida Geol. Survey Second Ann. Kept., pp. 21-173, 1909. Vaughan, T. W., A contribution to the peoIo;;ic history of the Floridian Plateau: Carnegie Inst. Washington Pub. i:!3, pp. 99-185, 1910; see also U. S. Geol. Survey Prof. Paper 71, pp. 741-745, 1912. of the formation, is a thin series of interbedded greenish sands and marls overlying the Oak Grove sand. It was named from Shoal River, in western Florida, and is not represented by a lithologic unit at Alinn Bluff. The following section was taken at the point where the fossil plants were collected, near the lower end of the bluff and in the immediate vicinity of the section measured by Dall.- It is deemed worthy of reproduction because it differs in certain particulars from Dall's sec- tion. Still other sections from different parts of the bluff are given by Sellards and Gunter.' Section at Alum Bluff, Fla. Pleistocene (?): rect. Light-colored ferruginous, rather loose sands. 9 Hard reddish clay 2 Variegated reddish and yellowish ferruginous sands 0-5 Miocene: Ohoctawhatchee marl: Dark-gray p\Titiferousclay, moreor less carbo- naceous but scarcely meriting the term lig- nitic given to it by Langdon, as no lignite or plant fossils were obser\ed in it. Traces of in\-ertebrate fossils, for the most part un- determinable, were obser\ed in places. The pyritiferous character of the clay gives it an alum-like taste, wliich accounts for the name of the bluff . Approximate thickness. 25 Bluish (when unweathered) fossiliferous clay marl of irregular tliickness, canying Muli- nia cowjesta, Ecphora quadricostata, Turri- tella rariahilis, and other species; much o.xi- dized in its upper portion, in whirh the fos- sils are represented liy poor casts, ow'iug to the solution of the sliell substance 15-30 Erosion unconformity. ~ Dall, W. H., and Stanley-Brown, J., Cenozoic geology aloug the Apalachicola River: Geol. Soc. America Bull., vol. 5, ji. 157, 1S94 s Sellards, E. H., and Gunter, Herman, Florida Geol. Survey Second Ann. Rept., pp. 275, 276, 1909. 41 42 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY, 1916. Oligocene: Alum Bluff formation: Feet. Cross-bedded laminated sands with clay lam- inae and thin distorted clay lenses of small extent. The sands are locally tliicker bedded and argillaceous, especially in the upper part, where they are in places packed witli tlie distorted detached rays of aSabal- like palm. In tlie lower part they are more evenly bedded anrown lignitic clay 2 Unconformity (local?). 4. Yellowish or gray argillaceous fine sand or sandy clay, grading into underlying beds 10 .5. Similar materials of greenish color, weathering yellowish, irregularly bedded and carrying iMge numbers of poorly preserved plant re- mains, mostly palm rays; grades into under- lying beds; thickness about 4 G. Yellowish or greenish much-jointed sandy clay.. 6 _Track level ISfeet above low water of Leaf River Figure 8. — Section of Hattiesburg clay near Raglan, Miss. Alum Bluff is a most mteresting spot to the physiogi-apher, because of the striking contrast in topogi-aphy between the east and west banks of the river; to the geologist, because it is a classic locality for the so-called Old Miocene and represents also the oldest knowai southern outcrop of the Chesapeake Miocene; and to the botanist, because in the wooded and damp ravines along the bluff are to be foimd two isolated gymnosperms, both relics of bygone floras, Taxus finridana NuttaU and Tumion taxifolium. (Arnott) Greene. Alum Bluff is of particular interest to the paleobotanist, be- cause it furnishes a representation of an upper- Numbers correspond to those given in the section in the text. The plants in the sands at both the localities mentioned above are few and very friable. Collections could not be made in the usual way, and as my improvised method jjroved very satisfactory it is worth describing. A small excavation in the loose, slumped sand was lined with a large piece of burlap. This was partly filled with plaster. While the plaster was still soft the sandy specimen was placed in it, the face of the specimen being left about on the level of the plaster. After the plaster had set the face was thickly covered with cot- ton, over which the free ends of the burlap were wrapjx-d and tied firmly. Every speci- PHYSICAL CONDITIONS AND AGE INDICATED BV FLORA OF ALUM BLUFF FORMATION. 43 men treated in this way, some of which were 10 inches in diameter, survived the rough handling on the river boat and the freight transportation to Baltimore. Fossil, plants from Alum Bluff have been mentioned by Langdon,' Foerste,- and Dall.^ A few fragments of Sabalites were collected by Dall, but no systematic collections were made until I visited this outcrop in 1910 in comjiany with E. H. Sellards, State geologist of Florida. CHARACTER OF THE FLORA. The flora is limited to 13 described species, although fragments of other species are present, and I observed but did not succeed in collecting a palmately veined Ficus, leaves of Gj'minda or Xanthoxylum, and pods resembling those of the existing Gleditsia aqiiatica Marsh. The determined species comprise a spot fun- gus (Pestalozzites), a very abundant fan palm (Sabalites), and 11 species of dicotyledons, in- cluding an elm, breadfruit, buckthorn, cam- phor, satinwood, ironwood, and persimmon. There are 1 1 genera in 9 families and 8 orders. These famihes are the elni (Uhnaceae), mulberry (Moracese), pisonia (Nj^ctaginacese), senna (Csesalpiniacese), i"ue (Rutaceie), buckthorn (Rhamnace^), laurel (Laurace£e), sapodilla (Sapotacese), and ebony (Ebenacese). The families Lauraces? and Sapotaceis are each rep- resented by two species; the remaining seven families have each a single species. By far the most abundant form is the palm, broken stipes and detached rays of which are thickly crowded in the sands in places. Four of the plants are ordinarily considered strictly tropi- cal— the breadfruit (Ai-tocarpus), brasiletto (Cffisalpinia), Nectandra, and satinwood (Fa- gara). The genera Pisonia and Cinnamomum are commonly considered tropical, but Pisonia reaches the keys of southern Florida in the ex- isting flora and the camphor tree (Cinnamo- mum) ranges northward to southern Japan and to the rain forests of southwestern China, while Cinnamomum camphora Linne is hardy in cidtivation around Tallahassee, Fla., and is ' Lansdon, D. W., Some Florida Miocene: Am. Jour. Sci.,3d ser.,vol. 38, p. 322, 1889; Geolog>- of the Coastal Plain of Alabama, p. 373, 1S94, 2 Foerste, A. F., Studies on the Chipola Miocene of Bainbridge, Ga.. and of Alum Blufl, Fla.: Am. Jour. Sci., 3d ser., vol. 46, pp. 244-254, 1893; Fossil palmettos in Florida: Hot. Gaz., vol. 19, p. 37, 1894. ' Dall, W. H., and Stanley-Brown, Joseph, Cenozoic geology along the Apalachicola River: Geol. Soc. America Bull., vol. 5, pp. 147-170, 1894. not uncommon as an escape from cultivation in woods and thickets throughout peninsular Florida, being freely seeded by birds. In fact, soU, humidity, and the length of the growing season seem to govern the extension of the tropical flora into the temperate zones to a much greater extent than actual extreme -tem- peratures, the existing floras of both south- eastern Asia and southeastern North iVmerica showing many parallel examples of such ex- tensions. The Sabalites is represented in the existing flora by the genus Sabal Adanson, which is now confined to America. It consists of eight coastal or stream-border shrubs and trees, five of which are confined to the West Indies, Mexico, and Venezuela; one is confined to peninsular Florida, and two range northward along the Atlantic coast to the Carolinas. The large mass of frayed and tangled rays and stipes of Sabalites in a matrix of sandy alluvium both at Raglan, Miss., and Alum Blirff, Fla., suggest that at the time these de- posits were laid down the shores were low and were densely clothed with palmetto "swamps" or brakes. The genus Ulmus, although it has tropical aUies, is in the existing flora a strictly north- temperate form having about 16 widely dis- tributed species. The family Rhamnaceas is mostly tropical, but several of the genera extend into the Tem- perate Zone, and Rhamnus, in particular, is mostly extratropical in the North Temperate Zone. There are about 75 existing species, and of the dozen North American forms several range northward to Canada and British America and only 1 ranges as far south as Florida. The genus Bumelia has about a score or more of existing species, ranging from Brazil northward through Central America and the West Indies to the United States, where two species are found as far north as Virginia and Illinois. The genus Diosp}TOs belongs to a large family that is mostly tropical in its distribution. Of the more than 200 existing species our common Diospyros i^irginiana Linne is found as far north as southern New York aiid New England. The genus is represented in south- ern Europe, and there are several species in eastern Asia. Moreover, of the 100 or more 44 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY, 1916. fossil species that have been described many occur in associations that are obviously tem- perate in character. To summarize the climatic conditions indi- cated by the flora, they are those C)f a tropical flora becoming replaced by a temperate flora, namely, subtropical or very warm temperate. These conditions are obviously different from those indicated by the Chipola marine faxma, which comprises over 400 knowni species, mostly Mollusca, and indicates a shallow sea (maximum depth not over 20 fathoms) and, according to Vaughan,' strictly tropical temperatures — that is, the bottom temperature of the water did not go below 70° F. dm-ing the year. As has already been indicated, the flora em- braces a number of tropical types which, as is shown by the existing flora, are legitimately to be expected to extend more or less beyond the equatorial belt in areas where rainfall is abun- dant and where extremes of low temperature are absent. Associated with these are forms like Ulmus, whose modern representatives are prevailingly north temperate but some of which extend into boreal regions. Still other of the Alum Bluff forms, the Rhamnus, Bumelia, and Diospyros, represent families wliich are mainly tropical in the existing flora but which as repre- sented by these genera have extended over large areas of the warmer parts of the Temperate Zone. It seems to me that among existing plant assemblages the Alum Bluff flora represents three types of plant associations. One corre- sponds in a general way to the "low hammock"' of present-day peninsular Florida, a type inter- mediate between the true hammock and the swamp type. A second plant assemblage indi- cated is that of the low-lying semiswamp pal- metto-brake type, along with some forms of the sandy strand, such as Pisonia, Cjesalpinia, and Fagara. In other words, this flora would find a congenial habitat at the present time in the delta of Apalacliicola River or almost any- where along the coast of peninsular Florida. Although so much less extensive, the Alum Bluff flora is somcwliat less tropical in its f acies than the flora of the Wilcox group and decid- edly less tropical than the floras of the Clai- borne, Jackson, Vicksburg, or Catahoula. On the other hand, it has not nearly the temperate facies of the flora of the overliving Chesapeake 'Vaughan. T. W., Carnegie Inst. Washington Pub. 133, p. 156, 1910. Miocene and also indicates much more humid conditions than the latter. This statement is based on the flora of the Calvert and a consid- eration of the abundant marine faunas of the Calvert, St. Marys, Yorktown, Duplin, Clioc- tawhatchee, and Jacksonville formations. The Alum Bluff flora may be considered to be the result of a reversal of the history of the present flora of peninsular Florida. That is to say, the present flora represents primarily a temperate flora receiving additions from the Tropics, whereas the Alum Bluff flora repre- sents an endemic tropical flora gradually becom- ing invaded by members of a temperate flora as a result of changing climatic conditions. AGE OF THE FLORA. The stratigraphic relations clearly indicate that the Alum. Bluff flora is younger than the Vicksburg Oligocene and older than the Chesa- peake Miocene. The faunal and floral evidence is equally conclusive. Not a single Alum Bluff plant is common to the Oligocene (Vickslnirg and Catahoula) floras of Mississippi, Louisiana, and Texas nor to the Chesapeake Miocene floras, which are, however, smaller than those from the Oligocene. Moreover, the facies of the Alum Bluff flora is decidecUy different from that of the floras of any of these horizons. There are no western United States or West Indian fossil floras for comparison, so that it remains to con- sider the probable European ecpiivalents of the Alum Bluff. As the Alum Bluff flora is obviously younger than the abundant European floras of the San- noisian (Lattorfian, Tongrian) and Stampian (Rupelian) and older than the exceedingly rich floras of the Helvetian and Tortonian (Vindo- bonian) , the only stages remaining are the Chat- tian (Kasselian), Aquitanian, and Burdigalian (Langhian , Mayencian ) . The Chattian (Fuchs, 1894) is the lower Aquitanian of Munier Chalmas and De Lappa- rent (1893). It is considered the equivalent of the Kasselian and in the Paris Basin is repre- sented, according to Lemoine, by the meulieres de Montmorency. According to Hang its in- vertebrate fauna is distinguished by the absence of Miocene types. The Aciuitanian (Mayer, 1857) now has the nari-ow limits assigned by DoUfus (1906-7). It marks the maximum regression of the Oligocene PHYSICAL CONDITIONS AND AGE INDICATED BV FLORA OF ALL'M BLUFF FORMATION. 45 soa due to the elevation of the PjTenees and the begiiniing of Alpine orogenesis. Marine waters invadetl the marginal coasts in Aquitaine, the Gironde, and Provence, in southern Spain, in Italy, and at various points on the south side of the eastern Alps; but the great bulk of the Atjuitanian sediments are those of lakes, swamps, and lagoons, with lignites and abun- dant and widespread mammals and plants. In Aquitaine the Tongrian is said to grade imperceptibly into the Aquitanian both litho- logically and faunally, and the latter passes into the BunUgalian in the same gradual man- ner. The floras of the Aquitanian arc like- wise transitional in character between Oligo- cene and Miocene. The marine faunas, how- ever, are said by Haug to contain only 4 per cent of Oligocene species and many Miocene species, but the foraminiferal genus Lepido- cyclina passes without modification from the Chattian into the Aquitanian. The Aqui- tanian has long been considered the uppermost stage of the Oligocene, although many paleon- tologists have pointed out the resemblances between the upper Aquitanian floras and faunas and those of the Burdigalian. In recent years, under the leadership of DoUfus and other French students, the Acpiitanian has been made the basal stage of the Miocene, although the question of its reference to the Oligocene or Miocene is vigorously disputed. The Burdigahan (Deperet, 1892) is closely related to the Aquitanian both faunally and florally. The marine faunas as well as the ter- restrial floras are said to indicate a shght lowering of temperatures since Aquitanian time. Tectonic changes had caused the dis- appearance of the broad lakes of the Aqui- tanian, and the Burdigalian materials comprise marine sediments on the southern and west- ern borders of the continent and a series of river, flood-plain, and swamp deposits (brown coal) in the region extending from France to Bohemia. It can not be said that the Alum Bluff flora offere conclusive evidence for detailed correla- tion, as it is too small. .Ul but two of the species are new and offer only indirect evi- dence. Of these new species the Ulmus is very similar to Ulmus longifolia Unger, of the Aquitanian of Bohemia, Germany, Styria, and France, and the Sapotacites is most like Sapotacites {Clirysophi/llum) sagorianum, from tlie Aquitanian of Sagor, in Camiola, described by Ettingshausen. The two species with an outside chstribution, Cinnainomum sclieuchztri and Diospi/ros hrachysepala, have both been identified in beds from a great va- riet}- of horizons in Europe. Although many of these identifications are of doubtful value, both species had a wid<^ range, geographic as well as geologic. Th(* original descriptions of both, by Alexander Braun, were based on material from the Tortonian of Oeningen in Baden. The two species are typically Oligocene- Miocene forms and both ani common and cliar- acteristic in the A(|uitanian and Burdigahan of Europe. It is thus apparent that the ^Vlum Bluff flora can be considered either Aquitanian or Burdi- galian, with a slight preponderance of the evi- dence in favor of the Aquitanian. Maury ' on the evidence of the molluscan fauna correlated the Chipola marl with the Aquitanian and con- sidered the Oak Grove fauna, as well as that of the sands at Alum Bluff above the Chipola, as ''transitional" between Aquitanian and Miocene. If subsecjuent paleozoologic studies corroborate Maury's work, there will be sub- stantial agreement between the floral and faunal evidence. If the Alum Bluff formation is of Aqui- tanian or Burdigalian age — and one or the other alternative seems certain — the more or less academic question is raised whether it shall be classed as Ohgocene or Miocene. Since the proposal of the term Oligocene by Beyrich in 1854 many have questioned its utihty or ultimate survival. Certainly there is but slight structural (diastrophic) evidence for placing the Oligocene-Miocene boundary in Europe between the Aquitanian and Burdi- gahan, and the marine faunas as well as the terrestrial floras and faunas show a gradual transition from the one stage to the other, so that the French paleontologists will probably be followed in tlieir contention that the Aqui- tanian should be placed in the Miocene. In considering the American application of the term Ohgocene, it may be noted that tliero appears to have been continui)us and uninter- rupted sedimentation in the Florida area from the deposition of tlu^ uiulerlying Chattahoochee formation into Alum Bluff time. Tliere also seems to have been a succession of minor earth movements during this period, and there was a greater influx of terrigenous materials into ' Maury, C. O., Bull. Am. Paleontology No. 15, 1902, 46 SHOETEE CONTRIBUTIONS TO GENERAL GEOLOGY, 1916. the Alum Bluff sea than into the Chattahoochee sea, a change in sedimentary character more probabh' due to an inland rise of the land which accelerated erosion than to the shal- lowing of the sea. It seems obvious that the Alum Bluff forma- tion as a whole is a predominantly shallow- water deposit of clays and sands and that the Chipola, Oak Grove, and Shoal River members are faunal zones contained in successive lentic- ular beds in the clays or sands. The faunules of these zones are closely related but show, according to Dall,' certain elements of tran- sition in the Oak Grove sand from the tropical Cliipola fauna to one indicatmg a shght lower- ing of the temperature. There is thus no structural (diastrophic) evi- dence for drawing the Ohgocene-Miocene boundary between the Chattahoochee and Alum Bluff formations, nor is there any floral or faunal evidence for such a bomidary. There is such a break between the Vicksburg and Apalachicola groups, and the Alum Bluff is separated by an erosion unconformity from the overlymg Choctawhatchee Miocene. It rests with invertebrate paleontology to deter- mine whether or not the whole of the Apa- lacliicola group shall be considered Miocene. Whatever may be the final verdict, it remains true that the flora preserved at iUum Bluff re- cords the last phase of sedimentation before the area emerged from the sea and that the most profound break in Tertiary sedimentation in the southeastern United States, emphasized equally by epeirogenic, faunal, and floral changes, was at the end of Apalachicola time — that is, it is represented Ijy the unconformity at the top of the Alum Bluff formation. SYSTEMATIC ACCOUNT OF THE FLORA. Class FUNGI. Order MELANCONIALES. Family MELANCONIACE.ffi. Genus PESTALOZZITES Berry. Pestalozzites sabalana Berry, n. sp. Plate VIII, figure 3; Plate IX, figure 9. Essential characters unknown. Foimd in- festing the leaves' of Sabalites apalachicolensis Berry in considerable abmidance and causing 1 Dali, W. H., Contributions to tlie Tertiary fauna of Florida: Wag- ner Free Inst. Sci. Trans., vol. 3, pt. 2, pp. 1574-1575, 1903. the formation of leaf spots. These spots arc of definite form and regular outline, small and circular at first, becoming larger with age and elongated parallel to the long axis of the ray, thus becoming elliptical or lenticular in outline. Maximum size observed, 1.5 centimeters in length and 0.5 centimeter in width. Average size, about 6 by 2 mnUmeters. The appearance of the infested leaves is well iUustrated in the figures, and they are scarcely to be distuiguished from numerous leaves of the existmg scrub palmettos, as for example, Serenoa serrulata (Michaux) Hooker, infested with the existing leaf-spot fungus Pestalozzia sp. Occurrence: Hattiesburg clay, Raglan, For- rest County, Miss.; collected by E. W. Berry. Alum Bluff formation. Alum Bluff, Liberty County, Fla.; collected by E. W. Beny. Collections: United States National Museum. Class ANGIOSPERM.ffi. Subclass MONOCOTYLEDONiE. Order AEECALES. Family PALMACEjE. Genus SABALITES Saporta. Sabalites apalachicolensis Berry, n. sp. Plate VIII, figures 1-.5; Plate IX, figure 9. Leaves of variable size, the maximum diam- eter estimated (from collected material) at about r20 centimeters. Rachis large, Imear, not enlarged at the base of the leaf, lenticular in cross section, with straight unarmed edges, continued for a short distance on the lower side of the leaf as a rapidly narrowed acumen which is only 3.5 centimeters in length in the small specimen figured; abruptly rounded-truncate at the base of the leaf on the upper side, where an inconspicuous ligulc is present. Rays numer- ous, 40 to 60 in number, carinate, linear- lanceolate in form, expanding to their middle, and free for about the uj)per third of their length, more or less curved at the base. Max- imum width observed about 4 centimeters; average about 2 centimeters. Midribs of rays not especially strong or prominent. Second- aries numerous, fine, longitudinal, parallel, largely immersed in the leaf substance, which must have been coriaceous. No transverse veinlets observed. Tliis species has a recorded range along the coast of the Gulf of Mexico from Florida to PHYSICAL COXDITIOXS AND AGE INDICATED BY FLORA OF ALUM BLUFF FORMATIOX. 47 oontrul Mississijipi und is jii pLu'cs !il)mi(liUit but ovcrvwhoro fragmentiirv. In the absi'uci' of fairly well jireservecl specimens showing the rachis and acumen it is impossible to distin- guish this form froni the AVilcox species Saha- lites graq/anus BeiTv or the Oligocene species SdhdlJlift virlshnrgeiifiiN Berry, so that possibly the latterspecies, represented byincomplete ma- terial, may be present in the Alum Bluff sands, although I consider this extremely doubtfid. SabaUtes graijanus differs from the present species in the expanded upjier end of the rachis, the longer and more gi-adually naiTOwed acu- men, and the more numerous rays, which are usualh- more conspicuously veined. Tlie fragments of rays which are so common in some of the deposits of the Vicksburg groujj and MhicJr are made the basis of Sahalites inclcs- hiirgensis are to be distinguished from 5. apa- lachicolensis chiefly l)y their much more promi- nent venation and longer and more slender acumen. The present species at its region of maxinuim occurrence at Alum Bluff is badly jj^infested with a leaf-spot fungus which I have described as Pestalozzifes sahalana Berry, n. sp. The foliage of jialms is abundant and well distributed throughout the deposits of the Wilcox, Claiborne, Vicksburg, and Apalachi- cola groups, indicating the abundance of plants of this type in southeastern North America during the Tertiary period. Few of these remains represent entire leaves, and in many places only fragments of rays are preserved. Palm leaves are notoriously difficidt of deter- mination, and the bidk of remains represent- ing flabellate fan palms with an acumen are referred by American students to the genus Sa])alites and by European students to Sabal. Two methods of specific differentiation are possible. Minor differences and stratigraphic position may l)e ignored, as in the case of the geographically and geologically wide-ranging Sahal major Unger of Eurasia, or minor differ- ences that also represent differences in geologic age may have considerable weight in specific differentiation. The latter is the method that I have found most useful from both the bio- logic and the geologic viewpoint. The extreme tropical climate of Vicksburg and Catahoida time, which is reflected in the abundant traces of palms found in deposits of Vicksburg or Catahoida age from Texas east- ward, continued througli the time of deposition 30841°— 16 2 of tlie Chipola marl member of the .Uum Bluff formation. Fortunately conditions were fav- oralile for petrification, and fragments of petri- fied ])alm tnmks, many of them of large size, are abundant throughout the area underlain by the Vicksburg or Catahoula deposits. Severn species of Palmoxylon have been de- scribed from these deposits. So far as I am aware no petrified palms have thus far been (>l)tained from the Alum Bluff sands, })ut palm foliage is very abundant in the leaf-l)earing lens at Alum Bhiff. I have never seen so great an abundance of stipes and rays of palms as occur at this outcrop. In places whole laj-ers consist of a mass of frayed and tangled rays in a matrix of sandy alluvium. They are usuaUy much macerated, and onl}- here and there can larger fragments of leaves be found. They are almost equally a])imdant in the similar mate- rials croppmg out near Ilagian, Miss. This suggests the presence near the coast in late Alum Bluff time of extensive palmetto swamps or brakes along the lower reaches of a sluggish river or estuary ovUj a few feet above mean water level — not flooded or true swamps, but subject to periodic overflow. Occurrence: Hattiesburg clay, Raglan, For- rest County, Miss, (common; collected by E. W. Berry) ; and near Chicoria, on Chickasaw- hay River, Wayne County, Miss, (collected by E. AY. Berry). Alimi Bluff formation, .Uuni Bluff, Liberty County, Fla. (collected by E. W. Berry), and Boynton Bluff, Choctawhatchee River, Fla. (collected by E. H. Sellards). Collections: United States Xationtjl Museum. Subclass BICOTYLEDON.S:. Order FAGALES. FamUy XJLMACE^. Genus ULMUS Linne. Ulmus floridana Berry, n. sp. Plale IX, figm-es .5-7. Leaves of medimu or small size, ovate-lan- ceolate in general outline, -tt-ith slightly ine- quilateral cuneate roundeedo- drome course and near then- ti])s give oil: one or • two outwardly directed tertiaries, which run to the teeth or to the smuses. This well-marked s^iecies of I'hnus apjjears to have been common during Almn Bluff time and constitutes the one strictly temperate element in the Alum Bluff' flora. Among ex- isting sjjecies it is most similar to JJlmus alata Michaux, which ranges from western I lorida northward to Vh'gmia and from Texas to Illinois. Like U. alata,' the fossil species was probably an inhabitant of rich alluvial swamp and stream borders, for it seems ])robable that the fossiliferous lens in the Alum Bluff sands is the result of stream action. Among fossil 'species, of which many have been described, it bears more or less resemblance to a number of widely scattered forms, especially because the lunits of variation of the foliage among ■different species of Ulmxis are not wide. The most similar fossil form is TJlmus longlfoUa, de- scribed by Unger ' and subsequently I'ecorded from the Aquitanian of Bohemia, Germany, Styria, and France, wliich is extremely close to the iVmerican species. The genus Ulmus may be distuiguished from the allied genus Caiiiinus by the usually more inequilateral leaves and by the tertiaries to the marginal sinuses. Occmrence: Ilattiesbm'g clay, Raglan, For- rest County, Miss, (collected by E. W. Beriy). Alum Bluff formation. Alum Bliiff, Liberty County, Lla. (collected by E. W. Berry). Collection: United States National Museum. Order URTICALES. Family MOKACE.5:. Genus ARTOCARPTJS Fdrster. Artocarpus lessigiana floridana Berry, n. var. Plate X, figures 5-7. Leaves of large size, rather smaller than the type, but very ])oorly preserved in the cur- rent-bedded sands. At least 25 centimeters in 1 Unger, Franz, Chloris protogaea, p. 101, pi. 26, flg. 5, 1S47. length and about 13 centimeters in maxunum width, thiLs more narrowly oblong in general outline than the tyjie. Pinnately S to 10 lobed; the lobes short, relatively broad and conical, acutely pointed and directed obliquely outwi.rd, separated by veiy naiTow sinuses. Midrib stout. Lateral jjrimaries stout, diverg- mg from the midrib at angles of about 45°, one to each lobe, terminating in its tip. Secondaries alternatmg with the ])rimaries, one to each sinus, the latter with the charac- teristic marghial hem. Tertir.rios mostly ob- solete. ^Vrcolation quadrangular where seen. Texture coriaceous. There was evidently con- siderable variation in outline, for the basal portion of the leaf sho^v^l in figm'e 6 has a cuneate base and must have had strongly ascendmg lobes. This fonn is based on very fragmentary specimens from the Alum Bluff sands. It may lie distinguished from Artocarpus lessigiana by its relatively nairower form and its finer venation. The latter comes from the Wilcox group, a much earlier horizon, and has not been found in the inter\-ening inten'al, repre- sented by the Claiborne, Jackson, Vicksbui^, and Catahoida deposits. I have a feeling that the Alum Bluff' material represents a new species, but I hesitate to set up a species on such fragmentary material, which is. however, the best obtauiable by the most careful col- lecting. Remains of Ai'tocarj^us have been found m Europe, Greenland, and the United States. In this country we have the Laramie-Denver tyjie, which is considered the parent stock of the present variety. This t,ype and two other species are represented in the late Wil- cox of Louisiana and Ai'kansas by excellent material. There is a Fort Union species and another Eocene form on the Pacific coast. In Europe several species range from the LTpper Cretaceous to the Pliocene. There are about 40 existing species of Arto- carpus, all endemic in the southeastern Asiatic region, ranging from Ceylon tlu-oughout Malay- sia to Cliina and represented by cultivated forms in all tropical countries. They appar- ently did not become extinct in North America until the interval between the deposition of the Alum Bluff formation and the migi-ation of the Chesapeake Miocene fauna into the Florida region, a migration indicating a lowering of PHYSICAL CONDITIONS AND AGE INDICATED BY FLORA OF ALUM BLUFF FORMATION. 49 temppraturos entirely sufficient to explain the extinction of Artocurpus on the mainland, but not offering u satisfactory explanation of its failure to survive in more southern latitudes. Occurrence: Alum Bluff formation, Alum Bluff, Liberty County, Fla. (collected by E. W. BeiTv). Collection; United States National Museum. Order CHENOPODIALES. FamUy NYCTAGINACE.«;. Genus PISONIA Linne. Pisonia apalachicolensis Berry, n. sp. Plate X, figure 1. Leaves of relatively large size for this genus, obovate in general outline, with a ])road apex, which is evenly rounded or bluntly pointed, and a gradually n.irrowed, sharply cuneate base. Length about 5.75 centimeters; maximum width, above the middle, about 2.4 centimeters. Margins entire. Texture coriaceous. Petiole short and stout, about 6 milhmeters in length. Midrib stout but immersed in the thick leaf substance, curved. Secondaries entirely im- mersed and obsolete. The modern species of Pisonia are numerous; they occur cliiefly in the Tropics in both hemi- spheres and are largely coastal types. They are abundantly developed in Central America and tropical South America, and several species occur in the AVest Lidies and Antilles. About 15 fossil species have been described, the ear- liest recorded being from the Upper Cretaceous of both America and Europe. The lower Eocene of southeastern North America has fur- nished two well-marked species, there is a spe- cies said to be represented by both leaves and fruit in the basal Eocene of the Rocky Moun- tain area, a fourth species occurs in the Clai- borne group, and a filth has been found in the Jackson. Perhaps the most similar fossil form, based on both leaves and fruits, is Pisonia eocenica Ettingshausen,' from the lignites of Haering, in Tyrol, which is upper Eocene or lower Ohgocene in age, occurring also as early as the Lutetian of England. Among existing species several are close to the present form. Pisonia lon/jifolia Sargent, of the beaches and shores of lagoons from the ' Ettingshausen, Constantin, Die tertiiire Flora von Baring in Tirol, p. 43, pi. 11, figs. 1-22, 1853. Florida keys through the West Indies to Brazil, is very similar to Pisonia apalachicolensis, though only about two-tliirds as large. Occurrence; Alum Bluff formation, Alum Bluff, Liberty County, Fla. (collect(>(l by E. W. Berry). Collection: I'lii'od Slates National Museum. Order ROSALES. Family CSSALPINIACEjE. Genus CiESALPINIA Linnfi. Csesalpinia sellardsi Berry, n. sp. Plate IX, fi-mros 1, 2. Leaflets small, sessile, inequilateral, ellipti- cal in outUne, 7.5 millimeters in length and 4.2 miUimeters in greatest width. Apex broadly rounded. Base strongly inequilater- ally truncated, one margin ascending and the opposite margin subauriculate. Midrib slen- der. Secondaries two or more, slender, ascend- ing, camptodrome. Margins entire. This species, which is obviously new, is clearly the leaf of some shrub or tree of the family Csesalpiniaceae. Though only a few specimens were observed, this can not be considered as indicating scarcity, for the plant material at Alum Bluff is all rather meager and poorly preserved. It seems probable that the present species was a mem~ber of the strand flora, the strand being the habitat preferred by a number of species of this genus in the modern tropical American flora. The fossil species may be compared with a number of existing species, with which it shows a very close agreement. Among pre- viously described fossil species it is very close to what has been identified by Schenk - as Ciesalpitna townslundi Heer, from the Stampian of Sieblos, originally described by Heer ^ from the Aquitanian and Tortonian of Switzerland and Baden and identified by Geylcr from the Messinian of Sicily. There is a very notable display of Mimo- sacese and Ciesalpiniacene in the flora of the WUcox group, and these elements probably continued to lie prominent throughout the remainder of the Eocene and the Oligocene, 2 Schenk, August, in Zittel, K. A., Handbuch der Palaeontologie, Atith. 2, Palaeophytologie, p. 700, fig. 369 (4), ISiK). ' Heer, Oswald, Flora tertiaria Helvetife, vol. 3. dI. 137, figs. 26-37, 1M9. 50 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY, 1916. although the record is much less completely preserved in the post- Wilcox sediments. The present species is probably but one of many related forms that inhabited southeastern North America during the deposition of the Alum Bluff formation. It is more inequilateral than any of the known Tertiary species from tliis general region, although it might perhaps be considered a descendant of Caesalpimtes finsonensis Berry, a somewhat smaller, less obUque, and more coriaceous form, from the sands of middle Wilcox age in Madison County, Tenn. It is named for Dr. E. H. Sellards, State geologist of Florida, who visited this celebrated locality with me and helped collect the fossil plants. Occurrence: Alum Bluff formation. Alum Bluff, Fla. (coUected by E. W. Berry). Collection: United States National Museum. Order GEKANIALES. Family RUTACE.a;. Genus FAGAKA Linng. Fagara apalachicolensis Berry, n. sp. Plate IX, figure 2. Leaves compound. Leaflets sessile, of me- dium size for this genus, elliptical in general outline, with a broadly rounded apex and a broadly pointed base. Length about 3 centi- meters; maximum width, above the middle, about L75 centimeters. Margins entire. Tex- ture coriaceous. Midrib stout, somewhat flexuous. Secondaries four or five, suboppo- site to alternate pairs, irregularly spaced, diverging from the midril) at angles varying from 35° to 50'^ rather straight in their courses, abruptly camptodrome close to the margins. Tertiaries tliin, more or less im- mersed, forming small quadrangular or polyg- onal meshes. The genus Fagara, which contains more than 150 living species of slirubs and trees, is cosmopoUtan in tropical and subtropical coun- tries. A few forms more properly referable to Xanthoxylum range for considerable distances in the Temperate Zone, particularly in south- eastern North America. Fossil forms based on foliage are usually confused with the closely allied genus Xanthoxylum Linne. Leaf- lets referable to Fagara are not uncommon throughout our southern Tertiary deposits, several different forms from the Eocene and Oligocenc having been described. The genus appears to have been especially prominent in the torrid flora of the Vicksburg group, and tlie leaflets are very abundant in the clays of that age. The present species is clearly unlike any of these, being especially different in its broad apex and nari;owed base, anil might readily be confused with the small entire leaves of some live oak. The scarcity of leaf remains in the Alum Bluff formation renders any remarks on the relative abundance of their described flora without much significance. Occurrence: Alum Bluff formation, Alum Bluff, Liberty County, Fla. (coUected by E. W. Berry), and Boynton Bluff", Choctawhatchee River, Fla. (collected by E. H. Sellards). Collection: Laiited States National Museum. Order RHAMNALES. Family BHAMNACE.ffi. Genus KHAMNTTS Linng. Rhamnus apalachicolensis Berry, n. sp. Plate IX, figure 8. Leaves rather large for this genus, broadly elUptical in generiil outline, with a broad and evenly rounded apex and base. Length about 8 centimeters; maximum width, in the middle part of the leaf, about 4.6 centimeters. Mar- gins entire, eveidy rounded. Texture subcoria- ceous. Midrib stout. Secondaries stout, four or five alternate pairs, diverging from the mid- rib at angles of about 55° to 60°, curving up- ward almost immediately in a broad, sweeping curve, becoming subparaUel with the lateral margins, along which they arch camptodromely. Tertiaries thin, closely set, subparaUel, mostly percurrent at right angles to the midrib. This large leaf is distinct from the numerous fossil species of Rhamnus previously described, although it resembles a number of them more or less closely. The genus has about 60 exist- ing species, widely distributeil in nearly all tem- perate and many tropical parts of the world and found on aU the continents except AustraUa. The genus Rhamnus is fairly prominent in Tertiary floras of southeastern North America, the Rhamnacere being also represented by forms of Zizyphus and Paliiu'us of tropical type. Six species from the Wilcox group have been described, but none are yet known from PHYSICAL CONDITIONS AND A(iK INDICATED BY FLORA OF ALUM BLUFF FORMATION. 51 the Claiborne, Jackson. Vicksbiirg, or Cata- houla. Occurrence: Alum BlufT formation, Alum Bluff, Liberty Countv, Flu. (collected by E. W. Berry). Collection: United States National Museum. Order THYMELEALES. Family LAURACE.ff: Genus NECTANDRA Roland. Nectandra apalachicolensis Berry, n. sp. Plate IX, figure 3. Leayes oblong-oyate in general outUne, broadest near the middle and tapering to the bluntly pointed apex and base. Length about 8 centimeters; maximum width, half\yay be- tween the apex and the base, about 2.5 centi- meters. Margins entire, eyenly curyed. Tex- ture coriaceous. Midrib stout. Secondaries about 10 subopposite pairs, diyerging from the midrib at wide angles, about 60°, rather straight in their coui-scs two-thirds . of the dis- tance to the margins, where they purye abruptly upward and form a succession of small camp- todrome arches along the margins. Tertiaries obsolete. The genus Nectantlra has about 70 existing species confined to tropical and subtropical America, several of which are practically iden- tical with this Alum Bluff species. There are numerous fossil species, the genus being well represented throughout the Eocene of south- eastern North America, especially in the sedi- ments of the Wilcox group, from which at least fiye species are knowm. It is represented in the Claiborne and Jackson but has not yet been found iii the Vicksburg or Catahoula. The present species is not particuhu'ly close to any of the described fossil forms. Occurrence: Alum Bluff formation, Alum Bluff, Liberty County, Fla. (collected by E. W. Berry). Collection: United States National Museum. Genus CINNAMOMUM Blume. Cinnatnomum scheuchzeri Heer. Plate X, figure -I. Cinnamomvm scheuchzeri. Heer, Flora tertiaria Helvetise, vol. 2, p. 85, pi. 91, figs. 4-22; pi. 92; pi. 93, figs. 1, 5, 1856. This species was described by Alexan,der Braun from both calyx and leayes, obtained in the Tortonian of Oenmgen, Baden. Typical ma- terial is common in the type area in both the Aquitanian and the Btu-digalian. Only the original tlescription of Heer, which is accom- panied ])y ample figures, is cited aboye, for sanguine students haye fancied that they had found this species at all horizons from the Up- per Cretaceous to the Pliocene, and at a very large number of localities throughout the Northern Hemisphere. Some of these identi- fications are undoubtedly correct, and the species certainly had a wide geographic and geologic range. Other identifications are un- questiona])ly erroneous, but it is impossible to sift the good from the bad without access to the original material, and I haye therefore not attempted to giye the synonymy or range. The Alum Bluff material appears to be iden- tical with a part of Heer's material from the type area, and it is also of the same age, so that I have no hesitation in identifying it with Heer's species. It denotes an ovate-lanceolate leaf, rather abruptly pointed at the extremi- ties, the base bemg broader than the apex. Length about 8 centimeters; maximuni width, in the middle part t)f the leaf, 2.7 centimeters. Margins entire. Texture coriaceous. Midrib stout; Lateral primaries one on each side, sub- opposite, suprabasilar, camptodrome. Second- aries thin, three or four camptodrome pairs in upper half of the leaf. Occurrence; Alum Bluff formation, Alum Bluff, Liberty County, Fla. (collected byE. W. Berry) . Collection; United States National Museum. ft Order EBENALES. Family SAPOTACE.a:. Genus BUMELIA Swartz. Bumelia apalachicolensis Berry, n. sp. Plate IX, figure 4. Leaves oblong-obovate in genert^l outlme, with an evenly roruuled apex and a narrowed cuneate base. Length about 4 centimeters; maximum width, above the middle of the leaf, about 1.9 centimeters. Margins entire, rather full. Texture coriaceous. Petiole short and stout, about 2.5 millimeters in length. Midrib stout, especially proximad, curved. Seconda- ries thin, numerous, 10 to 12 subopposite to alternate pairs, subparallel, at approximately regular intervals; they diverge from the mid- rib at angles of about 40° and are camptodrome 52 SHORTER CONTEIBVTIOXS TO GENERAL GEOLOGY, 1916. in the marginal region. Tertiaries immersed in the leaf substance. In the modern flora the genus Bumelia em- braces about 20 species of shrubs and mostly small trees, confined to the Western Hemis- phere, where they are distributed from the southern United States through the West In- dies, Mexico, and Central America to Brazil. Some of the species range northward to Vir- ginia and southern Illinois. They inhabit for the most part the strand, sandy soil near the coast, river bottoms, and the borders of swamps. Fossil species of Bumelia are numerous and the genus was probably cosmopolitan during the Tertiary — it was certainly common in the Eu- ropean area. It has been continuously repre- sented in southeastern North America since the Upper Cretaceous. Four lower Eocene species from this area have been described, one of which, Bumelia pscudotenax Berry, from the Wilcox group of northern Mississippi, is not luilike the Alum Bluff species but somewhat smaller and relatively narrower. Other spe- cies are present in the deposits of the Claiborne and Vicksburg groups. Compared with existing American species the present form is closer to the temperate than to the tropical species. It is intermediate between Bumelia tenax WiUdenow and B. lanuginosa Persoon and may stand in an ancestral rela- tionship to these modern forms. The former ranges along the coast from Cape Canaveral to North CaroUna and the latter from northern Florida along the Gulf coast and up the Mis- sissippi Valley to southern IlUnois and is abun- dant and of its largest size in the river bottoms of eastern Texas. Occurrence: Alum Bluff formation. Alum Bluff, Liberty County, Fla. (collected by E. W. Berry) . Collection: United States National Museum. Genus SAPOTACITES Ettingshausen. Sapotacites spatulatus Berry, n. sp. Plate X, figure 2. Leaves of medium size, obovate or spatulate in general outhne, with a broadly rounded apex, from which it narrows gradually with nearly straight lateral margins to the sharply cuneate base. Length about 7 centimeters; maximum width in the upper part of the leaf, about 2.75 centimeters. Margins entire. Tex- ture coriaceous. Petiolar portion missing. Midrib stout but more or less immersed. Secondaries obsolete by immersion. Ter- tiaries shown in microscopic preparations to form a very close meshed areolation. Sapotacites is a form genus for generically undifferentiated or undeterminable members of the family Sapotacete, and numerous species that range from the Upper Cretaceous through, the Tertiary have been described. It is pos- sible to refer many of these ancient species, such as the numerous forms in the flora of the Wdcox group, to Bumelia, Mimusops, Side- roxylon, Chrysophyllum, and other allied gen- era. The present form is much like a number of existing species of Mimusops as well as some forms of Bumeha. The family is chiefly trop- ical and subtropical. No fossil species are especially close to the present one, although it shows considerable resemblance to CTirysopliyUum sagorianum Et- tingshausen,' from the Aquitanian of Sagor, n Carniola. Occurrence: Alum Bluff formation. Alum Bluff, Liberty County, Fla. (collected by E. W. Berry). Collection: United States National Museum. Family EBENACEJE. Genus DIOSPYROS Linne Diospyros brachysepala Alex. Braun. Plate X, figure 3. Diospyros brachysepiilii. Alex. Braun, Die Tertiar-Flora von Oningen: Neues Jahrb.. 1845, p. 170. Diospyros brachysepala. Heer, Flora tertiaria Helvetise, vol. 3, p. 11, pi. 102, figs. 1-14; pi. 153, fig. 39b, 1859. Diospyros brachysepala. Friedrich, Boitrage zur Keniit- niss der Tertiarflora der Sachsen, pp. 63, 119, 126, 253, 255, pi. 6, fig. 1, 1883. Diospyros brachysepala. AVard, Types of the Laramie flora: U. S. Geol. Survey Bull. 37, p. 104, pi. 49, figs. 1, 2, 1887. This early described species has been re- corded from a large number of American and Eurasian locahties ranging in age from basal Eocene to Pliocene. It is improbable that a single species existed for so long a period, and I have therefore reduced the synonymy of ' Ettingshausen, Constantin, Die fossile Flora von Sagor in Krain, pt. 2, p. 14, pi., 12, flgs. 1&-21. 1877. PHYSICAL CONDITIONS AND AGE INDICATED BY FLORA OF ALUM BLUFF FOBMATIOX. 53 this species to a few representative citations.' Tlie species has been recorded by Lesquereux, Ward, and others from tlie lower Eocene of the Rocky Mountain area, as well as by Heer from West Greenland. I am not cer- tain that all these records are correct, for Diospyros hrachi/sepala is typicallj- an OUgo- cene-XIiocene form. It is found in the type area in both the Aquitanian and Burdigahan. The jVlum Bluff leaves, like most of the material referred to this species, are relatively small, about 5.5 centimeters in length and 2.8 centimeters in maximum width, midway be- tween the apex and the base, which are about equally pointed, the general outhne being al- 1 A much more complete sj-nonymy is given in my paper The lower Eocene floras of southeastern North America: t.". S. Geol. Suney Prof. Paper 91 (in press). most eUipticiJ. The midrib is stout and the secondaries tliin, numbering six or seven sub- opposite to alternate camptodrome pairs. The leaves here described are close to a number of examples figured by Heer from the type local- ity, but are relatively slightly broader than some of the forms referred to this species by different authors. In some respects they are very similar to Diospyros lamarrnsis Knowl- ton,- from Lamar (probabh' upper Miocene), in Yellowstone Park. OccmTonce: ^Uuiii Bluli formation, .Vlum Bluff, Liberty County, Fla. (collected by E. W. Berry). Collection: United States National Museum. « Knowlton, F. H., U. S. Geol. Surrey Mon. 32, pt. 2, p. 751, pi. 95, figs. 5, 6; pi. 96, fig. 4, 1899. PLATES. 55 U. S. GEOLOGICAL SURVTY PROFESSIONAL TAPER 98 PLATE VII A. PLANT-BEARING BEDS IN THE HATTIESBURG CLAY NEAR RAGLAN, FORREST COUNTY, MISS. n. PLANT-BEARING BEDS AT THE TYPE LOCALITY OF THE ALUM BLUFF FORMATION, ALUM BLUFF, APALACHICOLA RIVER, LIBERTY COUNTY, FLA. U. S. GEOLOGICAL SURVEY PROFESSIONAL PAPER 98 PLATE VIII FOSSIL PLANTS FROM THE ALUM BLUFF FORMATION AT ALUM BLUFF, FLA. PLATE VIII. Figures 1-5. Sabalites a palachicoleH»is Berry, n. sp., from the Alum Bluff formatinn. Alum Bluff, Liberty Count>', Fla. Figure^ 2 and 5 show prolongation of the racliis as a midrib on the under side of the leaf. Figure 3 shows fragment of a leaf infested with Pestalozzites sabalana Berry, n. sp. Fig^ire 4 shows the upper side of a leaf base and the termination of the rachis as a truncate ligule. 57 PLATE IX Figure 1. CTsalpinia selkirdsi'BeTr}-. n. sp. Figure la. The same, X 5. Figure 2. Fagara apalachicoUnsls Berry, n. sp. Figure 3. Nectandra apakicMcoknsis Berry, n. sp. Figure 4. Bumelia apalachicolensis Berrj', n. sp. Figures 5-7. Ulmus floridana Berry, n. sp. Figure 8. Rhamnus apahchicolensis Berry, n. sp. Figure 9. Sabalites apalachicolensis Berry, n. sp. Portion of a very large leat infesteJ with Pestalo^zites sabalana Berry, n. sp. 58 U. S. GEOLOGICAL SURVEY PROFESSIONAL PAPER 98 PLATE (X FOSSIL PLANTS FROM THE ALUM BLUFF FORMATION AT ALUM BLUFF, FLA. U. S. GEOLOGICAL SURVEY PROFESSIONAL PAPER ^8 PLATE X FOSSIL PLANTS FROM THE ALUM BLUFF FORMATION AT ALUM BLUFF, FLA. PLATE X. Figure 1. P!so)iia apdlarlucolfiisis Berrj^ n. sp. Figure 2. Sapotacitcs spatulatus Berry, n. sp. Figure 3. Dk.t,pyros hrachysepala Alexander Braun. Figure 4. Cinnamomum scheudizeri'S.eQX. Figures 5-7. Arlocarpus lessviixina fluridana Beriy, n. var. 59 o 3 SIR'S ;i(-,i