THE LATE NIAGARAN STRATA OF WEST TENNESSEE. By William F. Pate,Of Lebanon, Kentucky.andRay S. Bassler,Curator, Division of Invertebrate Paleontology, U. 8. National Museum. Probably nowhere in America is a more complete and betterexposed section of Niagaran strata than that shown along the Paleo-zoic geanticline of West Tennessee. This section is well exposedalong the Tennessee River and its tributaries, but the many finefossils that have made this classic ground foi; the paleontologist havecome almost entirely from the conspicuous white-capped, roundedhills or " glades " scattered through portions of Decatur, Perry,Wayne, and Hardin counties. A portion of this Niagaran sectioncontains rather thick beds of Avhite clays and clayey limestones,which weather into soils unfavorable to the growth of much vegeta-tion. Therefore, when these particular strata happen to occupy thetop of the hills or are exposed along their slopes, spaces almostentirely barren of vegetation occur in otherwise well-wooded areas,and because of their white clayey soil, such bare spots or glades, asthey are locally known, are visible for considerable distances. Thestrata whose vlisintegration jn-imarily gave rise to these spots areusually very fossiliferous. In process of erosion the fossils are leftbehind, so that in the course of time most of the glades become idealcollecting grounds. Indeed, fossils are sometimes so abundant onthe glades that specimens can be shoveled up literally by the bushelmeasure.Despite the fine fossils and good sections afforded by the area,comparatively little has been done on its paleontology, and but onewriter has published on its stratigraphy in any detail. The presentcontribution is based upon several short seasons of study and collect-ing by Mr. Pate, and upon a joint study of the rocks by both authorsduring two weeks of the summer of 1907. Much of the seniorProceedings U. S. National Museum, Vol. XXXIV?No. 1621. 407 408 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxiv.author's time was devoted to collecting the echinoderm fauna of thearea for Mr. Frank Springer, but, in tracing the various crinoidhorizons, an exceptional opportunity was afforded to study thestratigrajjhy in detail. This collecting resulted in the addition ofmany new species to the already considerable crinoid fauna and inthe discovery of material affording new data concerning thedescribed forms. These crinoids are at present being siudied by Mr.Springer, who expects to publish a special monograph upon them inthe near future. Besides the study of the stratigraphy with Mr.Pate, the junior author paid particular attention to the collection ofthe other classes of fossils, so, by a combination of efforts, it ishoped to add something to the knowledge of the region. Practicallyall of the echinoderms found on these various trips are a part of theSpringer collection, while all of the remaining specimens belong tothe U. S. National Museum.In 1847, Roemer visited this glade region and spent about fiveweeks collecting the fauna. The country Avas then very sparselysettled, and his efforts Avere, of necessity probably, confined to De-catur County, in the vicinity of Dixon Spring and Brownsport Land-ing, where he could secure living accommodations. Practically allthe species which in ISOO were described by him in '' Die SilurischeFauna des Westlichen Tennessee " were collected on the glades of thelocality mentioned. Since Roemer\s species have been procured fromthe same sections and localities b}^ subsequent collectors, there isnow no doubt regarding their geologic position. Troost, Sa'fford,Worthen, Wachsmuth, Eominger, and other paleontologists collectedon the glades repeatedly, but, aside from the publication of a newsjDecies now and then, nothing further has been done upon thepaleontology.The first account of the stratigraphy of these Silurian rocks wasgiven by Professor Saff'ord in 1861," and later in his Geology of Ten-nessee, in 18G9, where the Niagaran rocks are subdivided into twonearly equal portions, each about 100 feet thick, a lower or variegatedbed and an upper or sponge-bearing bed. The general section pre-sented by Professor Safford, taken from the tyjDe locality at Clifton,is as follows:(2) Meniscus limestone.(h) Sponge-bearing bed; gray, erinoidal. and argillaceous lime-stones, many of them glade-forming, highly fossiliferous,containing sparsely thin layers of chert 90 feet.(a) Variegated bed; gray, red, and mottled limestones, inter-stratifled; many layers argillaceous; Orthocerata abundantin its lower part. Clifton is located in part upon it 96 feet.Entire thickness of the formation 186 feet. ?Amer. Jour. Sci. (2), XXXI, pp. 206, 207. LA TE NIAGARAN STRA TA?PA TE AND BASSLER. 409The name Meniscus limestone is misleading, since the characteriz-ing sponge, Astrceospo7i(/ia meniscus, is found only in the upper orsponge-bearing bed. In 1876 this name was changed to the Cliftonlimestone by Safford and Killebrew in their Elementary Geology ofTennessee.In 1903, Dr. A. F. Foerste " published his valuable paper " Silurianand Devonian Limestones of Western Tennessee,'' by far the mostimportant contribution to this subject. Doctor Foerste's studies ledhim to recognize in Salford's Variegated beds the Clinton, Osgood,Laurel, and Waldron formations of the western flank of the Cincin-nati geanticline, and at the top of this division two new formations,the Lego limestones and Dixon beds. To Satford's upper member,or sponge-bearing bed, as exposed at Clifton and Brownsport Fur-nace, Foerste applied the new name Brownsport. Other new names,the Maddox limestone of early Niagaran age, the Glenkirk limestonefor a combination of the Laurel, Waldron, and Lego, when thesecould not be separated, and the Gant bed for an arenaceous limestonephase of a part of these late Niagaran rocks, were instituted. Doc-tor Foerste's article contains much valuable information, but hisarrangements of facts, particularly in the description of sections, issuch that it is difficult to assemble his evidence. We are in accordwith most of Foerste's work on the lower portion of the Niagaran,and our efforts in this paper are to present a detailed accoimt of thelate Niagaran, with particular reference to the Brownsport divi-sion. We have made free use of Doctor Foerste's work and wish toacknowledge our indebtedness to his publication.A study of the stratigraphy of this general area has shown numer-ous unconformities in strata which are essentially horizontal. Thegeologic range is from the lowest Trenton to the Mississippian, butthe main geologic divisions are seldom conformable. The lowestTrenton is succeeded by the uppermost Cincinnatian strata; theNiagaran rocks are followed by Helderbergian strata of New Scot-land age, and even in these larger divisions unconformities are notedbetween their individual members.Such minor unconformities are indicated in the table of sectionson page 428. It may also be noted that, so far as observed, no singlelocality affords a complete and continuous section of all the Niagaranstrata known to occur in the area. For example, at the type-locality,Clifton, in Wayne County, a considerable portion of the late Niag-aran is wanting; even at Decaturville, where probably the most com-plete section of the entire area may be seen, several members aremissing. It is, therefore, only by comparing section after sectionthat the complete succession may be determined. The compositesection offered on a subsequent page was thus compiled. * Journal ?f Geology, XI, pp. 554-715. 410 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. XXXIV.As noted before, our attention wa.s directed more particularly tothe Brownsport bed and succeeding strata, or, in general terms, thelate Niagaran rocks. These strata have furnished by far the mostof the fossils from the glade region, and reference by most authorsto the Niagaran fauna of Western Tennessee is not to the Cliftonlimestone as a whole, but to this particular division. Careful col-lecting and detailed stratigraphic work has convinced us that thepo.strDixon strata of Niagaran age, instead of being a bed of hetero-geneous clayey shale and limestone, referable to only one formation,the Brownsport, contains at least four divisions well marked faunallyand lithologically, and interesting furthermore because of their un-conformable development. Accepting the term Brownsport anddefining it as a group, we would propose in addition the followingnew formations: Beech River, Bob, Lobelville, and Decatur. Theseveral classifications proposed for these rocks may then be tabulatedas follows : Niagaran strata of West Tennessee. Saflord, 1869. N0.1621. LATE NIAGARAN STRATA?PATE AND BASSLER. 411 CLIB'TON, TENNESSEE.The section at Clifton is well exposed along the river bank, the low-est beds being shown north and south of the old cement mill, where afault brings the Middle Ordovician Hermitage limestone in contactwith the Dixon. Proceeding southward from this ])oint, the Dixonand overlying formations are well exposed ; northward along the riv^erbank and then along the slope of the hill back of Clifton, the entiresection can be made out in more or less detail. The Waldron, withits typical fauna, is Avell shown along the river bank about one-fourthof a mile south of the cement mill and also in the toAvn of Cliftonitself. The following section is a combination of these several ex-posures. In this, as well as in the following sections, the strata arearranged under the formations and subdivisions recognized by us,and described more in detail on succeeding pages.Sect ion lit Clifton, Win/nr Count i/, Tennessee.Gravel beds.Silurian (Niagaran).10. Beech River formation. Feet.c. Eucalyptocrinus zone. Dark blue shales with many chertylayers. Fossils abundant 36J&. Troostocrinus zone. Blue shales and limestones holdingTroostocri)ius and associated fossils and arranged asfollows : 5. Limestone layer ' 14. Blue shale 133. Compact limestone 22. Blue shale 181. Compact limestone in layers varying from 1 to Ginches in thickness 5a. Coccocrinus zone : 5. Blue shales with thin beds of nodular chertylimestone G*4. Blue shales with four or more well defined chertbands 1 to 2 inches thick 33. Blue shales free from chert, irregularly developed- 0-72. Grayish white crinoidal limestone with a purplishtinge when freshly broken 641. Blue shales weathering white upon exposure 29. Dixon formation.Red shales and clayey limestones with occasional layers ofwhite argillaceous limestones. Fistulipora hemisphericarather abundant 448. Lego limestone.h. Compact gray limestone with layers of reddish, moreargillaceous rock 14Ja. Compact, white, massive limestone 3147. Waldron shale.White indurated clay and white argillaceous limestoneswith the typical Waldron fauna, among the species being 412 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.xxxiv.Feet.Euc(i1i/pt(jcriints Diagnus, E. elrodi, 8tci)}ianocrhiiis genimi-formis, Dictyonclla reticulata, Meristiiia iiiaria, Honiwo-spira craj\ Nuclcosinra pisifonnis, Anat^tropliia internas-cens, Lcptaiia rhnnihoidiiUs. Rhipido)iicUa hyhrkla, Cal-lopora clrgantitht, etc 36. Laurel limestone. ^Massive, reddish purple argillaceous and subcrystalline lime-stone 255. Osgood limestone.Thin-bedded reddish argillaceous limestone holding Stephano-criiiiis os'(/(>od('iisi-s and t.vi?ical (UinjocrUmx ornatiifi 144. " Clinton " forma tion.<^White to light brown crystalline fossiliferous limestone withgreenish brown chert. This layer has yielded the foUow-hig fossils : Lcptcena rhomhoidaJis, Dalmanella elcgantula,riatystrophia daytoneii.sis, Orthis flahcUites, Cyclonema day-toiicnsis, Ilhrniis daytmtciD'iis, Calymene rogdcsi, Cyrtoccrasfiubcoiiiprcs.'^inn, Favosites faiH)sus, F. niagarensis, and Haly-sitcs ciitcnulatiiK 1Ordovician ( Cincinnatlan ) . ."5. Fernvale (Richmond) formation.h. Bluish shales with Rhijiichotrcma capax, Dinorthis suh-quadrata, D. proarita. and Rhoinhotrypa quadntta.Mannie clay of Foerste loi(/. Coarse-grained, cross-beddetl, light-colored jihosphaticlimestone with h'hymhofrona capa.r and Htrophomena])l(iiif)dorsata. Lei])ers limestone.'' 192. Arnheim ("Warren") formation.Coarsely crystalline phosphatic limestone weathering intoabundant chei't fragments. Dinorthh rctrorfta, Lrpta'narlioiiihoiddlis. and Rhynchotrcma drntluish-gray, compact argillaceous limestone layers two tofive feet thick, alternating with blue shales of equal thick-ness. 70+BROWNSPORT FURNACE. TENNESSEE.In the vicinity of Brownsport Furnace two well-marked glades ex-pose a fine section of the various beds belonging to the Beech River,Bob, and Lobelville formations. The red limestones and shales of theDixon may be seen at the base of the southern glade, while the young-est Paleozoic rocks of the section?the Devonian black shale?wereexposed in the hillside northwest of the furnace. " This name is provisionally employed for the western rocks referred to theClinton by Foerste and others.* Foerste, not Leipers formation, Hayes and Ulrich. NO. 1621. LA TE NIA aARAN STRA TA?PA TE AND BASSLER. 413 Section at Brownsport Furnace, Decatur County, Tennessee.Gravel hods.Devonian. Feet.("liattanooga shale (exposed alonj; hillside northwest of furnace) __ ?Silurian.0. Lobelville formation.Coral z(?ne. Yellow clays and aricillaceous limestones hold-ins ill! abundant coral fauna. Contains numerous speci-mens of several si)ecies each of Fi'.vozom 3Residual limestone bowlders holding Camurocrinus and Helder-bergiau brachiopods 5Silurian.Decatur limestone. Massive, white crinoidal limestone becomingyellow and shaly at the top 70Beech River formation.Eucalyptoerinus zone. Bluish shales becoming yellow towardthe top 20Troostocrinus zone. Fossilferous shales varying from bluethrough yellow to white in order, with a limestone band in themiddle part 24Coccocrinus zone. Shaly limestone in the upper part and whitishlimestone easily disintegrating into clay in the lower 25Dixon formation. I'uridish shales in lower part and red crinoidallimestone in upper 35Lego limestone. Grayish limestone with a few red layers 30Waldron shale. Purplish shales with an intermediate layer of whitelimestone. MimuJm icaJdroiiciiHifi and other Waldron fossils noted- 4Laurel limestone. Pale purple, massive limestone 25Osgood limestone. Massive white crinoidal limestone with a tend-ency to form cliffs when favorably exposed 10Ordovician.Aruheim ("Warren") formation. Coarsely crystalline phosphaticlimestone 2Hermitage (" Saitillo") formation. Argillaceous limestone and shalewith usual Lower Trenton fossils. This, the lowest rock of thevicinity, is exposed for a distance of a mile along Rushing Creekeast of Decaturville 50lady's bluff, perry county, TENNESSEE,This in one of the highest and most beantiful bhiffs along the Ten-nessee River, and furnishes an excellent section of the Niagaran andHelderbergian rocks. Lady's Blutf is on the east side of the Ten-nessee River, 1 mile below Mousetail and just below the mouth ofLick Creek, in Perry County. In this section, which follows, theabsence of the coral beds normally preceding the Decatur limestoneis noteworthy.i^cclion at Lady's Bluff, Perry County, Tennessee.Gravel beds.Devonian.Linden (Helderbergian) formation. Feet.h. Shales with numerous bryozoa and brachiopods 3a. Massive crinoidal limestone with Camaroerinus 4Silurian,Decatur limestone. Massive crinoidal limestone, gray, magnesian inupper part and more yellow in lower portion 63Lobelville formation.Coral zone wanting.Bryozoan zone.Light blue soft shale 2White argillaceous limestone 2 NO. 1621. LA TE NIA GARAN STRA TA?PA TE AND BASSLER. 415Feet.Blue ^hale with bryozoa and a few corals 1Purple shale and argillaceous limestone, containing manybryozoa and a few corals 17Bob formation.Conchidiuni zone. Xodular, cherty, crinoidal bine and firay lime-stone with Coucliidiuin in upper part esi)ecially. Other fossilsfew 15Dictyonella zone. Blue shales with rather numerous brachiopods_ 5Uncinulus zone. Grayish, cliff forming limestone breaking intolarge blocks 16Beech River formation.Eucalyptocrinns zone. Blue and yellow shales with EucnUji)tn-crinus and other characteristic fossils. At one point in the nearvicinity this bed has thickened to 40 feet 20Troostocrinus zone. Yellow shales and thin white limestone 24Coccocriuus zone. Only topmost layer exposedMOUSETAIL, PERRY COUNTY, TENNESSEE.The local variability of the beds comprised in the late Niagaranis brought out in a comparison of the Lady's Bluff section and thatat Mousetail, 1 mile south. The absence of the Bob formation inthe Mousetail section is particularly noteworthy.Section at Mousetail, Perry County, Tennessee.Decatur limestone. F*^6t.Massive gray and white coarsely crystalline limestone with few fossils_ . SLobelville formation.b. Coral zone. Yellowish clays and thin limestones with numerousspecimens of Farosites. HeJioUtes, and other corals 8a. Bryozoan zone. Soft shales holding many bryozoa and a few coralsof the genera Favosites and HelioUtes.Bluish white shales 2Purplish shales 7Beech River formation.c. Eucalyptocrinns zone. Yellowish shales with crinoids and otherfossils characteristic of thisrfiorizon 276. Troostocruius zone. Thin limestone and shales with T. rein-wanltl, etc 45a. Coccocriuus zone 15Dixon formation.Red shales. Top only exposed ?In the remaining sections studied by us the differences in lithologyand in fauna were so slight that it was deemed best to arrange thesein tabular form. In this table the bottom and top of each section isindicated by the lowest and uppermost figures, or X, while blankspaces within these limits show the absence of formations. Thevarious thicknesses are given in feet when accurately determined,but by X when only noted. 416 PROCEEDINGS OF THE NATIONAL MUSEVM. VOL. XXXIV. IN(N NO. 1621. LATE NIAGARAN STRATA?PA TE AND BASSLER. 417 POST-l)I.\ON FORMATIONS.Alon^ the western flank of the Cincinnati axis the Sihirian rocksfollowing the Waldron shale are known as the Lonisville limestone.In West Tennessee the same interval is occnpied by several limestoneand shale formations wliich have been named by Foerste, in ascend-ing order, the Lego limestone, Dixon bed, Brownsport bed, and Gantbed. The Brownsport and Gant beds are the special subjects of thispaper.The term Brownsport Avas loosely defined. In fact, tlie name coiddhardly be considered as defined at all, since practically the only de-scription of these strata given by the author of the name is to theeffect that the name Brownsport bed is applied to the richly fossil-iferous section overlying the Dixon red clay and containing the faunastudied by Roemer during his visit to Decatur County, Teimessee,In this sense the name would have reference mainly to the crinoid,brachiopod, and coral beds, so conspicuously shown in the section atBrownsport Furnace. Apparently the name was instituted only forconvenience of reference, and not as a formational name. In thesame article (p. 570) rocks now known to be equivalent to the brachi-opod beds of the Brownsport Furnace section are named the Gantbed. The recognition of either or both of these names as fornuitionsis thus made (piestionable, and we were in doubt whether (1) to ap-ply the name Brownsport as a group term to the post-Dixon Niag-aran strata as exposed at Brownsport Furnace, (2) to extend the termso as to include all the Niagaran rocks following the Dixon, or (8) torestrict and redefine the formation entirely. After a considerationof all the sections, the first course seemed the wisest.In the post-Dixon Niagaran interval four well-marked divisionscan, Ave believe, be recognized. The lowest is a series of ratherwidely distributed, very fossiliferous Avhite clays and thin-beddedlimestones furnishing most of the crinoids afforded by the glades;argillaceous limestones and shales, (piite similar lithologically to theDixon but of more local distribution, succeed these white strata, andare overlaid in turn by blue shales bearing great numbers of brachio-pods; just above the brachiopod beds are blue shales and cherty lime-stones in which the great coral fauna so Avell deA^eloped at Louisville,Kentucky, is likewise abundantly represented. In exposures of thesestrata in Tennessee the corals are sometimes so numerous that theground is entirely hidden by the specimens. These three formationsconstitute the BroAvnsport group as here recognized. Finally, theNiagaran of West Tennessee is closed by a massiA^e Avhite, coarselycrystalline limestone, 70 or more feet thick, Avhich seems to be of con-siderable extent, especially in the northern half of the area. 418 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxiv.BEECH RIVER FORMATION.As indicated previously, the lowest beds of the Brownsport groupfurnish most of the crinoids of the late Niagaran of West Tennes-see. To this formation the name Beech River is applied from theconspicuous exposures along Beech River in Decatur County. Theweathering of these beds gives rise to the most conspicuous of theWest Tennessee glades. Their strata are almost exclusively lineshales of a gray, yellow, or bluish color, but weathering into a Avhiteclay. These shales are generally present where their proper place inthe section is exposed; indeed, they form the most widely spread di-vision of the late Niagaran. Their fauna is also the best known ofany of the Niagaran divisions, probably because most of the Avell-known crinoids and blastoids from West Tennessee are derived fromthem. C'orcocrinus hacca, TroostocrhuiH reimrardtt^ and Eucalypto-crinits miUtgaude are particularly abundant, and each was found tobe characteristic of a certain zone. Numerous outcrops of theseshales may be seen in the general area between Perryville and Clifton,but particularly fine exposures are found at Brownsport Furnace,thence northward to Perryville, and at numerous localities alongBeech River. The Brownsport Furnace section has been given on apreceding page; the sections along Beech River are essentially thesame, but the best exposed and most clearly defined section of theBeech River formation is found at the large glade about 2 milessouth of Perryville.Srciidii tiro uiilc>> soutli of I'criijvlUc, Decatur County, Tennessee. Feet.Decatur limestone. In large boulders covering formation below ?Beech River formation.Eucalyptocrinus zone.Fine bluisli fossiliferous shales in the lower part and yellowshales increasing in their yellow color and fossiliferous charactertoward the top. Euvatijptocrinu.s niiUiganac is not uncommon,W'hile Lam])tcroerinus tennesseensis. Calticrinns ramifer, Eu-cali/ptocrinus rcntrieosus, Marsupiooinus tennesseensis, andmany other species of the same class are occcasionally found.Troostocrinus is a rare fossil and the specimens are larger andmore slender than the typical 2\ reinwardti of the beds below_ 20Troostocrinus zone.Yellowish shales with numerous typical Troostocrinus rcin-tcardti and Caryocrinus bulbulus less common 10Gray, shaly limestone forming a bench on the glade 1-3Very fossiliferous, bluish, and whitish shales, becoming tingedwith yellow toward the top 11Coccocrinus zone.Rather shaly limestone, much more resisting than shales aboveand below. This limestone forms a well-marked bench sepa-rating the Coccocrinus zone below and the Troostocrinus zoneabove 3 NO. 1621. LA TE NIA aARAN STRA TA?PA TE AND BASSLER. 419Feet.White and gray shaly limestones and shales which rapidly disin-tegrate into a white clay npon exposure. Voccocrinus baccafound throughout tliis division but most abundant near themiddle 19Gray, shaly, rather unfossiliferous limestones 5Hard white limestone with a purplish tinge npon fresh fracture_ 1White or greenish-white clay containing numerous dwarfedspecimens of Bilobitcs, I'icctambointcs, /'j.socn/(M.s, etc 2Dixon formation. lied and ])nri)le shales and shaly limestone ?Astrceospongia Tneniseits is a very abundant and characteristicfossil of the Beech River, aUhoupacc. rocks not seen ^ S(1) Argillaceons limestone at the foot of the hill, which may be referableto the Meniscus formation 5The list (column C) mentioned in the section was of brachiopodsabout equally divided between the Niagaran and " Lower Helder-berg '' species. The bed was, therefore, apparently transitional andof considerable interest in an area where the Niagaran and Helder-bergian Avere elsewhere unconformable.Doctor Foerste included this locality in his studies and found thata considerable stratigraphic interval separated layer 3 of ProfessorSafford's section from the Hardin sandstone (layer 4). He alsoproposed that " for i)urposes of discussion it will be found convenientto apply the name Gant limestone to the coarse sandy limestoneimder discussion [layer 3], and the term Gant hed to the top ofSilurian section, including the Gant limestone at its base." At theGant locality he found the " Gant limestone " to contain S-pirifer saf-fordl, Dlrtyonella (jihljosa^ Xudeospira concentrica, Wilsonia sa/foirliUnci/u/hhs stricklandi, and Schuchertella {Orthothetes) suhplanus.At other localities in the vicinity the " Gant limestone " was found tocontain Mcristina inaria-roemeri and Gypidida ivemeri^ in additionto the brachiopods already mentioned, and beneath this limestonewere found whitish clays and soft limestone containing sponges,brachiopods, and crinoids characteristic of the Glade exposures ofthe Beech River formation.The following section, which has been compiled from Foerste'sremarks upon this subject, is rather characteristic of his " Gant bed "in apparently the only area where it has hitherto been recognized.Section on hillside just northeast of Martin's mill, Tennessee.Bob formation. ' ^^?*-Layer C. Much weathered and poorly exposed limestone with Asty-lospongia praemorsa, Caryomanon stellatim-suleatuni, Meristinaniaria-roemeri, and Wilsonia saffordi 14 NO. 1621. LA TE NIA aARAN STRA TA?PA TE AND BASSLER. 421Feet.Layer R. Bluisli limestone, partly fine grained and i?artly crlnoidal,with Wilsonia saffordi, Unvinlus striclclandi, Hchuchertella ,sub-planus, Gypidula roemeri, Camarotoechia neglccta, Nucleospiraconcenirica, and Meristina maria-roemeri 9Layer A. Sandy " Gaut limestone " forming a projecting ledge 5Beech River formation.Whitish clays and soft?limestone, with sponges, braciiiopods, andcrinoids 57Dixon formation. Ued shale and shaly limestone.P'oerste did not recognize his " Gant bed '" in any but a small areaalong Indian Creek in the vicinity of Martin's mill and the Ganthomestead. We have found limestones and shales, carrying greatnumbers of brachiopods of the species listed by Foerste, succeedingthe Beech River formation and extending from a point just south ofDecaturville continuously eastward and southward through the areadrained by the Tennessee River and Indian Creek. At BrownsportFurnace these brachiopod beds are 25 feet in thickness, and it isstrange that the similarity of their fauna to that of the " Gant lime-stone " was not noticed hitherto. One-half mile below Bob, on thewest bank of the Tennessee River, the red limestones and shales aresucceeded by about 4:2 feet of limestones and shales abounding inbrachiopods. The section here is as follows:Section one-Uulf mile below Bob, Decatur Vounty, Tennessee. Feet.Lobelville formation ?Bob formation.(/) Hard grayish clay at top, soft yellow clay in middle part, andsoft blue clay at the base with brachiopod fauna listed below(under (c) ) 12(e) Hard buff limestone, blue upon fresh exposure 1-2(d) Soft blue clay with brachioimd fauna of bed below 2(c) Yellow argillaceous limestone layers, two to three inches thick,alternating with soft yellow clay. Brachiopods numerous, Gypi-dula roemeri, Meristina maria-roemeri, Nucleospira concentrica,Dictyonella gibbosa, Uncinulus tennessecnsis, and Wilsonia saf-fordi, being particularly abundant 14(b) Rather massive white limestone in 8 to 12 inch courses. Un-cinulus stricklandi not uncommon 10(a) Red limestones and shales to water's edge. Large Uncinulusstricklandi quite abundant 20As noted in the above section, all of these brachiopod bearing strataare grouped together in the Bob formation, this name being selectedon account of the excellent exposure at Bob Landing. Decatur County,Tennessee.The Bob formation is likewise found at a number of points northof the general area just described, and here the limestone part of thestrata seems better developed. At Lady's Bluff, 16 feet of limestoneProc. N. M. vol. xxxiv?08 28 422 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.xxxiv.and 5 of shale were noted in the section. It is possible that the orig-inal distribution of the formation was equivalent to that of the under-lying Beech River, and that erosion preceding the advance of thesea in which the succeeding formation?the most widespread of all?was laid down, has removed these strata, particularly in the morenorthern areas. *Although we might have adopted the term " Gant formation " forthese limestones and shales Avith an abundant and characteristic brach-iopod fauna, Ave came to the conclusion that the name had better notbe employed. At present there is scarcely any exposure at the originalGant localit3% and in a iew years vegetation Avill have completelycovered the rocks here. The place is found Avith great difficulty,and, moreoA^er, Gant no longer liA^es there. The term is not a geo-graphical one, and, in addition, the entire section is not developedthere. An excellent exposure of these strata occurs about 1 milefrom SAvift on the SAvalloAV Blutf road Avhere the shales and lime-stones hold fine specimens of the brachiopods. Here the limestonelayers, usually present at the base of the formation, haA^e disinte-grated in shaly material and the large Uncinulus stricklandi noted inthis particular bed occur free.Typical '' Gant '' limestone* and clays are also Avell exposed on theAvest side of Buffalo River in Perry County, at BeardstoAvn. Thesection along the Tennessee River in the vicinity of Bob is by farthe most complete and conspicuous, and it seemed most fitting, asstated above, to apply this new name to the formation.The upi^ermost division of the Bob formation is of nodular, chertydimestone or hard shale in Avhich brachiopods of the genus Con-vhldhim are the most characteristic fossils. On the east side of theTennessee RiA^er this zone, to Avhich Ave have applied the name Con-chidiuni, is Avell shoAvn in the Lady's Bluff section. Avhile Avest of theriver good exposures are found at BroAvnsport Furnace.LOBELVILLE FORMATION.The third formation here recognized is characterized paleontolog-ically by a fauna of corals so abundant in species and specimens thatthe identification of the beds is attended with little difficulty. Thesecorals are most abundant in the upper half of the formation, andAvherever their strata are Avell exposed many fine specimens may behad. Ilalysitefi eatenidatus., so abundant at Louisville, Kentucky, inthe late Niagaran, is represented in the West Tennessee strata,Avith the exception of the Clinton, only at this horizon. The sameoccurrence in the late Niagaran only, holds true for almost all ofthe folloAving species, likcAvise A^er}^ characteristic of the Louisvillecoral bed: Alveolites loiiisvillensis^ A. niagarensis, Amplexus shu-mardi, Cladopora complanata, C. reticulata, Coenites verticillata, NO. 1621. LATE NIAGARAN STRATA?PATE AND BASSLER. 423Favosites cristatus, F . cristatus-major^ F. discus^ F. farosiis, F . louis-inllensis^ F. nkigarensis, F. spongilla^ Ileliolites subtubulata^ II. inter- .stinctus, Plasniopora follis^ Calceola tennesseensis, Theeia major.! andT. minor. Short Creek, near Linden, Tennessee, is a famous localityfor these corals, and most of the West Tennessee specimens in paleon-tological collections were procured here. The strata, however, areexposed in Perry County from Linden to a point about 2 miles northof Lobelville, and, as the two faunal divisions which we recognize inthe formation are well shown in the vicinity of the latter place, thename Lobelville is proposed for this formation. West of the Ten-nessee River the Lobelville formation is widely distributed, but is notso continuously exposed. Its presence in the Brownsport Furnacesection has been indicated on a previous page, and other localitieswhere the formation is well shown, especially in the southern half ofDecatur County, might be named.Red, purple, and blue shales, holding many bryozoa and a numberof the corals noted above, follow the Conchidium limestone of the Bobformation and are succeeded by the strata in Avhich the coral faunais so well developed. These strata, for which we employ the sub-division Bryozoan zone, seem to be but 9 feet thick in the Mousetailsection. At Lady's Bluff, where a more detailed section was made(page 426) , a thickness of 22 feet was noted. Peeler's Pond, a locality1 mile northwest of Whitfield, in Hickman Comity, shows a sectionof red shales at the bottom closely resembling the Dixon lithologi-cally, and of blue shales at the top. These red shales are withoutdoubt not Dixon, but are the bryozoan beds noted at Mousetail andWebb's Mill. The bryozoa of these beds are mainly of new species.Several species of Fistulipora and a new Nicholsonella are particu-larly abundant. The occurrence of these red strata at the base of theLobelville formation, as well as similar red rocks at the same positionin the Bob formation, is believed to be very significant in indicatingprevious land conditions. , tSevtioa at Peeler's Pond, Hickman County, Tennessee.Devonian. ^*'*'*-Chattanooga black shale ?Silurian.Decatur limestone.Massive wliite limestone 6Lobelville formation.Coral zone. Shales and thin bedded limestone with an abun-dance of Favosites, etc 27Bryozoan zone.Soft blue shale with many bryozoa and a few corals 2.3Red clayey shales, greatly resembling Dixon formation, butcontaining characteristic Lobelville bryozoa and corals 8-1-Bottom not exposed. 424 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxiv.A section very similar to the above is exposed at the Webb or RiseMill, just south of Linden. The formations and thicknesses exposedhere are as follows:Section (It Webb or Rise Mill, near Linden, Tennessee. Feet.Devonian.Cbattauooj^a black shale ?Silurian.Decatur limestone 16Lobelville formation.Coral zone 40Bryozoan zone.Blue sbales with bryozoa and corals 18Red clay sbales witb base not observed 11A columnar section of this locality was given by Foerste " in Avhichthe basal red shales are referred to the Dixon. In his text this ref-erence is made doubtful, and he notes that the exposures aroundLinden are peculiar in not displaying the great wealth of brachiopodsand sponges characteristic of more southern exposures. If ourstudies are correct, these peculiarities are readily explained by theposition of the strata above the brachiopod (Bob) and sponge(Beech River) beds. To us the most striking feature of this andother localities near Linden is the great resemblance of the coral bedsto the late Niagaran section at Louisville, Kentucky. ? The u^Dper zone of the Lobelville formation is so well charac-terized by its prolific coral fauna that no difficulty should be experi-enced in its recognition. The strata are usually wdiitish shales soprolific in fossils that their disintegration upon hillsides leaves theground fairly covered by the specimens. A particularh^ fine exampleof a glade formed by the weathering of these coral beds is the Grave-yard Glade, Ih miles west of Bob. The section here extends fromprobably the top of the Beech River formation to the Decatur lime-stone occurring in bowlders on the top of the hill, but the corals fromthe coral beds have so covered the underlying strata that the succes-sion could not be determined with accuracy in the time available. Thecorals, determined from a preliminary study of the fauna in thishorizon at the Graveyard Glade, are listed beloAV. Every one of thesespecies occurs in the uppermost Niagaran strata at Louisville, Ken-tucky, and Ave do not know of a clearer case of equivalent rocks,judged by faunal evidence. ? Journal of Geology, XI, 1903, p. 578. NO. 1621. LATE NIAGARAN STRATA?PA TE AND BASSLER. 425 Corals at (irannard (lUieds ( l'vl)urii beds of Foerste) are crowded with fossils ot ^ewScotland a"e ' Stnutnpora issa. Favimttx conlrus, CaUoporaperelegans,Eridotrypa corticosa, and Lioclema cethdosum being particularlyabundant. ^^^'MassivTwhite, coarsely crystalline, crinoidal and magnesian limestone becoming yellow and slightly shaly at th?top. Lobelville formation. ^""'' YeHow ^cfavs and thin clayey limestone crowded with corals.Several .species of FavosUe><, Heliolites, Theeia, Cysttphyllum,Halysites, and Plasmopora are very abundant. "'^'whTte'to" blue' shales holding a few corals and many bryozoa inthe upper part; red to purple shales and argillaceous limestonewith the same fauna making up the lower division. 428 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxiv.Composite section of Ordovician-Devonian rocks of west Tennessee?Coutinuecl. .1621. LATE NIAGARAN STRATA-PATE AND BASSLER. 429Compimtc section of Ordoriciun-Deroman rocks of ircst Tennessee?Continued. g 1 Columnarsection(Dixon). Description. ^"^Compact gray t.. white argillaceous and mibcrysialline limestone withfew fossils. ^^'"^WhitlS'rated clay and argillaceous limestone, holding the typicalWaldron fauna. ^""MaS;%"nk to reddish purple argillaceous and crystalline lime-stone. '^::^S<<;i^ 70'+ Osgood Itoestone. argillaceous limestone containing Stephanocri-,Z osgoodensl^ PU^ocfinus c,emmiMmls, and typical Caryoennusornatus. " ^ Whi?e"to?S toown crystalline, fossiliferous limestones with greenishbrown chert. '''Tlulsh?es'inupper half, coarse grained, crossbedded, light colorednhosDlvitk' limesone in lower. Rhynchotrema eapax Dinortlns sh6-SadmS /A ^-o