DESCRIPTIONS OF SOME MAMMALIAN AND FISH REMAINSFROM FLORIDA OF PROBABLY PLEISTOCENE AGE.
Oliver P. Hay.Associate of the Carnegie Institution of Washington.
There are few of our States which give promise of furnishing moreimportant contributions to our knowledge of the vertebrate animalsof the Pleistocene than Florida. Already the list of species has be-come a long one and additions are constantly being made to it.Materials belonging to five species are described below. Two ofthese are beheved to be hitherto imnamed.ELEPHAS IMPERATOR Leidy.Plate 26, fig. 1.In 1889* Leidy described and figured a left ramus of the lower jawof an elephant which had been foimd by Mr. J. F. LeBaron, some-where along Peace Creek, probably not far from Arcadia, and whichLeidy identified as belonging to ElejjJias columhi. Leidy' s figurepresents a view of the worn surface of the tooth, which he recognizedas being the hindermost molar. He stated that there were twelveridges present and that these appeared to be the complete numberentering into the constitution of the tooth. Eight of these weresaid to occupy a space of 6.4 inches. Inasmuch as the tooth wasburied m the bone nearly to its summit, the thickness of the plateswas taken on the grinding surface.This jaw is in the United States National Museum, and has thecatalogue number 183. Recently the writer obtained permission toexpose the Ungual face of the tooth, and the result is shown on plate26 (fig. 1). Near their bases the space occupied by four plates isabout 95 mm. There can be no doubt that the jaw belonged to anindividual of Elephas imperator. The normal number of plates inthe last tooth of E. columhi is about 24. Had there ever been somany plates present the tooth would have had an enormous length.With 18 plates it was sufficiently large. Certainly some plates,about six, had been lost through usage.
> Trans. Wagner Free Inst. Sci., vol. 2, p. 23, pi. 8, flg. 2.Proceedings u. S. National Museum, Vol. 56?No. 2291. 103
104 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 56.TfflNOBADISTES, new genus.A new genus of ground sloths, most closely related to GnatJiopsisLeidy Type, T, segnis, a new species described below. Based on anastragalus, in the front half of the upper surface of which there is adeep ligamentous fossa opening forward; the lower face also crossedfrom front to rear by a ligamentous fossa.THINOBADISTES SEGNIS, new species.Plate 27, figs. 1, 2.Type specimen. An astragalus, No. 3335, of the U. S. NationalMuseum.Type locality.?WilUston, Florida.Type formation.?Pleistocene.Characters.?Those of the genus.In the United States National Museum there is a left astragalus ofa large ground sloth which is recorded as having been collected bythe United States Geological Survey in 1887, in Levy county,Florida. The catalogue number is 3335, and the bone is recorded aMylodon Jiarlani. It seems probable that the collector of the speci-men was Mr. J. B. Hatcher; and it is quite certain that it was foundat "Mixon's bone bed," near Williston, where many other fossilshave been secured.It appears that this bone had been studied by Doctor Leidy, forthere is writing on it in his chirography ; but he has not indicated onit any generic or specific name.Supposing that the bone belonged to Mylodon liarlani, the astraga-lus of which appears to be known only from Harlan's brief descrip-tion and poor figure,* the writer first compared it with that of Mylodonrohustus, as described and figured by Richard Owen.^ It soon be-came evident that the Florida bone was quite different from thecorrespondmg one of the South American species. On plate 27,figures 1, 2, are presented two views of the bone from Florida. Fromfigure 1 it will be seen that there is on the upper face a deep fossaextending from the front of the bone to its center. A large part ofthis fossa is occupied by a rough surface for hgamentous attachment,the apex of which is midway between the front end of the bone andthe hinder border of the articulation for the tioia. In all directionsaway from the fossa just mentioned the surface for articulation withthe tibia is strongly convex. In Mylodon rohustus the correspondingfossa is evidently much shorter and shallower, and the surface forthe tibia is much flatter and apparently even concave posteriorly.The greatest differences are seen, however, on the lower surface of
? Amer. Joum. Sci., vol. 44, 1843, p. 78, pi. 1, fig. 16.s Descr. skel. Mylodon rohustus, 1842, pp. 117, 131, pis. 21-23.
NO. 2291. MAMMALIAN AND FISH REMAINS FROM FLORIDA?HAT. 105the bone. Owen writes^ that the anterior and inferior surface of theastragalus of Mylodon rohustus is occupied by one extensive elongatedarticular surface adapted to the calcaneum, cuboides, and navicu-lare, and his figure shows that this is true. On the other hand, inthe bone from Florida, this surface is completely divided by a deeprough furrow for ligaments. It will be noted, too, that the out-lines of the two bones as seen from below are very different.It might be supposed that the astragalus here described is that ofMegdlonyx; but this bone was described by Leidy,^ and only a glanceat Leidy's figures is needed to convince one that the Florida bonecan not belong to that genus.Owen^ described and figured an astragalus which had been broughtfrom South America and which he thought belonged possibly toMegalonyx. This was afterwards made by Leidy * the type of a newgenus and species, GnatJiopsis oweni. When the Florida bone iscompared with Owen's figures here reproduced (pi. 27, figs. 3, 4)there are to be seen close resemblances. It might not be far out ofthe way to refer the astragalus from Florida to a second species ofGnathopsis, but a careful examination shows differences that seemto indicate a distinct but closely related genus. On the uppersurface of the bone figured by Owen there was certainly no suchdeep fossa for a process of the tibia and for a ligament as is seen inthe Florida bone. Nor was the surface for the tibia as convex asit was in the bone here described. Again, as seen from below, therewas in the astragalus of Gnathopsis a deep and wide notch in theanterior border at the end of the ligamentous groove, as if this andthe upper one joined across the border of the bone. In the Floridabone the anterior surface for articulation with the calcaneum extendsnearly to the inner border of the bone; in Gnathopsis oweni it ismuch shorter. Believing that such differences in as characteristica bone as the astragalus is among the ground sloths, indicate otherimportant differences in the skeleton, the name TMnohadistes isproposed for the genus, the species to be known as Thinohadistessegnis. (Derivations, Ols sand; ^ahcT-qs, a walker; segnis sluggish.)The following measurements in millimeters have been made onthe astragalus here described:Extreme length of astragalus 108Width from summit of tuberosity for tibia to border between thefibular and the calcaneal surfaces 97Length of surface for tibia 76Width of surface for tibia 70Height and length of surface for fibula 38Length of posterior articular surface for calcaneum 69Width of posterior articular surface for calcaneum 38
I Descr.skel. Mylodon rohustus, "p. 118.
? Smiths. Contrib. Know!., vol. 7, art. 5, p. 40, pi. 12, figs. 7-10.
' Descr.skel. Mylodon rohustus, p. 132, pi. 23, figs. 3, 4.
* Smiths. Contrib. Knowl., vol. 7, p. 41.
106 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 50.The posterior surface for the calcaneum is concave along its greaterdiameter; slightly convex along the shorter. The surface for thefibula is nearly plane in its upper part, but convex from front to rearin its lower half. The articular surface for the navicular is concave,but not deeply so. The surface for the cuboid is convex,TRUCIFEHS FLORIDANUS (Leidy).Plate 28, figs. 1-3.In 1889* Leidy described a skull of a saber-tooth tiger to whichhe gave the name Machairodus jioridanus. This had been securedby Mr. Joseph Willcox, in a limestone quarry at Ocala, Florida.From the same quarry had been obtained other remains which arereferred to Equus leidyi, Bison, sp. indet., Odocoileus sp. indet., Dasy-pus sp. indet., Sylvilagus sp. indet., Procamelus minor, and Elephascolumbi? All of these indicate that the deposits belong to thePleistocene.From the skull described by Leidy all the teeth were missing; butthere were present the alveoli for the upper incisors, the greatcanine, the third premolar, and the carnassial. This skull wasfigured in a later paper.^In the Eighth Annual Report of the Florida Geological Survey,on plate 29, figure 8, Dr. E. H. Sellards figured an upper carnassialpremolar which had been found at Vero, Florida, in the stratumknown in the literature of that locality as No. 2. This he referred(p. 152) to Smilodon. Recently, through the kindness of DoctorSellards, the writer has been permitted to examine the tooth inquestion. By comparing the figure of this tooth here presented(pi. 28, figs. 1, 2) with that of Leidy's Trucifelis fatalis* it will beseen that there is between them a close resemblance. It wiU benecessary first of all to determine whether or not the Vero toothbelongs to T. fatalis. The following measurements enable us tomake comparisons, those of T. fatalis being computed from Leidy'sdescription'^. In the second column under each tooth is given theratio of each dimension to the length of the tooth.Dimensions of upper camassials.
Trucifdisfdtalktype. Vero tooth.
Ijcngth of crownWidth at inner buttressHeight of principal cusp....Height of anterior lobeHeight of front of rear lobe
.
32.815.618.714.613.5
100415744.;41
10042.56547.547.5
1 Proc. Acad. Nat. Sol. Phila., p. 29.? Sellards, 8th Ann. Rep. Fla. Geol. Surv.. p. 103.
' Trans. Wagner Free Inst. Sei.. vol. 2, pi. 3, fig. 1.
? Ext. Mamm. Fauna Dak.. Neb., pi. 28, figs. 10, 11.6 Idem, p. 367.
NO. 2291. MAMMALIAN AND FISH REMAINS FROM FLORIDA?HAY. 107It will be seen at once that the Vero tooth was probably that of alarger species than T. fatalis and that the crown is everywhere higherin proportion to its length. As may be seen from comparing thefigures of the two teeth, the width of the principal cusp at its base isone-half of its height, while that of T. fatalis is relatively consider-ably wider. The two teeth agi-ee in having the protocone absentand in having the anterior lobe divided into two parts. It appearscertain that the Vero tooth does not belong to T. fatalis. The latterwas found at Sour Lake, in Hardin County, Texas.Through the courtesy of Mr. John G. Rothermel, director of theWagner Free Institute, the writer has been able to examine theupper jaw of Leidy's t^-pe of Machairodus floridanus. The carnas-sial possessed in front two roots, of which the inner was somewhatreduced in size and pushed backward, as in the Vero tooth, to nearlyopposite the interval between the anterior outer root and the greathinder root. It is quite certain that the anterior lobe of the toothwas much larger than it is in the lion and the tiger. The alveolushas a length of 37 mm. ; that part for the hinder root is 24 mm. long.The tooth was therefore only slightly larger than the Vero tooth;and there appears to be no reason why the latter can not be withmuch certainty referred to Leidy's species.In the deposit atVero which furnished the carnassial Doctor Sellardsfound a part of a great canine tooth Avhich belonged to some one ofthe Machairodontinae. The fragment (pi. 28, fig] 3) is 67 mm. long.Probably nearly 25 mm. of the distal extemity is gone. The upperend does not reach the base of the crown. In the Ocala skuU thesocket for the canine measures 40 mm. in length fore and aft and itswidth is 20 mm. At its upper end the fragment from Vero has awidth for and aft of 30 mm. and a thickness of 13 mm. If the frontand rear borders of this tooth are continued until the distance betweenthem is 40 mm. and the distal extemity is restored, a tooth is indi-cated whose crown was about 1 1 mm. long. Both borders are acute,more especially the hinder one, which is knife-like. The anterioredge is smooth, but the hinder one is obsoletely crenulated. Thetooth is quite different from that of Barnum Brown's Smilodontopsisconard%.' In the latter the base of the fragment has the same fore andaft diameter as does the Vero tooth. At a distance of 55 mm. fromthis, toward the tip, the fore and aft diameter is 19 mm.; in the Verospecimen, only 16 mm. In Smilodontopsis coiiardi both edges arecrenulated. In Cope's Smilodon gracilis => the powerful canine main-tains weU its breadth as the tip is approached; and both edges arefree from denticles (Cope). For comparison there is figured here(pi. 28, fig. 4) a right canine tooth evidently belonging to Dinolastis
1 Mem. Amer. Mus. Nat. Hist., vol. 9. p. 190. pi. 19.? Journ. .\cad. Nat. Sci. Phila., vol. 11, pi. 20, fig. 1.
108 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 56.serus Cope. It was found in a cave in the northern part of BexarCounty, Texas. The tooth belongs to the Scientific Society of SanAntonio. It was a far smaller tooth than that of figure 3,From the close resemblance existing between the carnassial foundat Vero and here referred to Leidy's Machairodus jioridanus and thatof Trucifelis fatalis found at Natchez, it is certain that both belongto the same genus. For this genus the writer accepts at present thename Trucifelis. We shaU have, therefore, the two species, Truci-felis atrox and T. jioridanus.FELIS VERONIS, new species.Plate 28, figs. 5-7.When the author was at Vero, in October, 1917, he found along thedrainage canal, a short distance above the railroad bridge, in the bedof sand known as No. 2, an upper left fourth premolar of a large tiger-like animal. Views of this tooth are here presented (pi. 28, figs.5-7). On comparing it with the corresponding teeth of the tigerand of the jaguar {F. imraguensis, No. 4128 U. S. Nat. Mus.), andwith those of the machairodonts, it can not be doubted that its pos-sessor belonged to a species of Felis. In order to facilitate comparisonthe following measurements are presented, being those of the speci-men in question, the same tooth of Felis tigris, and of the large SouthAmerican jaguar, Felis paraguensis HoUister.Measurements of carnassials of Felis.
Length of the crownWidth of crown at protoconeWidth of crown between the main and the hinder cuspsHeight of anterior lobeHeight of main cuspHeight of rear hinder lobeWidth of the main cusp
Felis from
NO. 2291. MAMMALIAN AND FISH REMAINS FROM FLORIDA?HAY. 109recent tiger or the fossil one. Naturally, the fossil is very distinctfrom the jaguar because of its greater size.From Natchez, Mississippi, Leidy described Felis atrox ^ which wasbased on a lower jaw with teeth. Inasmuch as the lower carnassial is31.2 mm. long, while that of the existing tiger above-mentioned isonly 23 mm., it is evident that F. atrox had upper carnassials whichwere about 44 mm. long. It was therefore a much larger animalthan the Vero cat. Felis augustus,^ besides belonging to the Arikareeof the Tertiary, differs in various ways from the Vero specimen.Felis JiiJlianus Cope belongs to the Blanco Pliocene and is based ona canine tooth and some foot bones; so that it can not be comparedwith the animal here described. Felis imperialis, of the Pleistoceneof California, had a second molar about 25.5 mm. long and was,therefore, a larger animal than that from Vero. According to Cope ^the upper carnassial of Fells inexpedata has a length of 24 mm., beingthus considerably smaller than that of the Vero animal.Inasmuch as this large felid found at Vero appears to have beenhitherto unknown, it is proposed to introduce it under the nameFelis veronis. TRICHECHUS ANTIQUUS Leidy?Plate 26, tigs. 2, 3.In the collection of the National Museum is a part of the lowerjaw of a manatee (Cat. No. 2522) v/hich is labeled as having beenfound with the other fossils of the Alachua clays, in Levy County.However, the writer finds no reference to this genus in any of the listsof materials collected in the Alachua clays; the fossil has an appear-ance different from most of the other fossils of those clays; and thereis attached to it an oyster shell, showing that it had lain in saltwater. Doctor Sellards informs the writer that he has never seenany marine fossils that have -been found in the Alachua clays. It ishence probable that the bone was found somewhere else in Florida.Leidy reported * fragments of ribs of supposed Tricheclius antiquusfrom Peace creek, and Sellards ^ included T. manatus among the fossilsfound in Withlacoochee river. No mention is found of the discoveryof a lower jaw at any place.The jaw in question appears to be well fossilized and it is heavy;so that it evidently belongs to either the Pleistocene or to somelate Tertiary deposit.The jaw lacks both ascending rami and all of the teeth. It evi-dently belonged to a species of Trichechus, but not to T. manatus.The individual possessing it appears to have had a size somewhatless than that of a manatee whose basilar length is 356 mm. The
1 Trans. Amer. Philos. Soc, vol. 10, 1853, p. 319, pi. 34. < Trans. Wagner Inst., vol. 2, p. 27.2 Leidy, Ext. Vert. Fauna, etc., pi. 7. sgth Ann. Rep. Fla. Geol. Surv., p. 104.
' Journ. Acad. Nat. Sci. Phila., vol. 11, p. 248.
110 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.50.distance from the front of the symphysis to the rise of the ascendingramus was close to 180 mm; in the jaw of the existing manatee withwhich it is compared this dimension is 195 mm. The length of thesymphysis is relatively the same as in the manatee. Its greatestheight is 68 mm.; in the manatee, 83 mm. In the latter animal theupper half of the hinder face of the symphysis forms a concavity;this does not exist in the fossil jaw. The surface which in life wasoccupied by the horny plate is relatively much shorter than in themanatee, being only 60 mm. long; in the manatee, 80 mm. In themanatee the inner face of the horizontal ramus is flat or even con-cave; this does not seem to have been the case in the fossil jaw.The height of the jaw was evidently less than in the existing manatee,being apparently only 48 mm. at the middle of the length; whereasin the only slightly larger manatee jaw the height is 60 ram. Theinferior dental canal is considerably larger than in the manatee, itsdiameter being 17 mm. Moreover, its outer face is open backwar?lto about the position of the third or fourth tooth.Judging from what remains of the sockets of the teeth the latterhad a length somewhat greater than in the existing manatee. Threeof these sockets occupy a line 45 mm. long; in the manatee used forcomparison, 41 mm. The lower teeth appear to have been widerthan those of the manatee, but of this one can not be certain.Leidy described an upper tooth of a manatee which bears thename Trichechus antiquus, and which was found at Charleston,South Carolina. The fore and aft diameter of the tooth v/as about20 mm.; that of the existing manatee is about 12.5 mm. T. an-tiquus was evidently a much larger animal. Its lov\^er teeth musthave had a length of about 24 mm. Evidently the jaw supposed tohave been found at Williston belonged to a considerably smallerindividual, perhaps to a smaller species, than the one which fur-nished Leidy's type. In the various species belonging to the genusTrichechus there is a continuous succession of teeth which are pro-duced at the rear of the jaw and which move forward. These in-crease in both length and width as the animal grows. Hence thetooth described by Leidy may have belonged to a very large speci-men of the same species as that to which the jaw belonged which isabove described. For that reason the jaw is referred provisionallyto Trichechus antiquus; but it may, vnih equal probability, have be-longed to an undescribed species.ATRACTOSTEUS LAPIDOSUS, new species.Plate 26, fig. 4; plate 28, fig. 8.In the United States National Museum are a right opercular boneand some scales of a fresh-water gar which are labeled as havingbeen found by L. C. Johnson, in 1885, in the "Mixon bone bed," in
NO. 2291. MAMMALIAN AND FISH REMAINS FROM FLORIDA?HAY. 1 1
1
Levy County, Florida. This place is near the present town ofWilliston. These gar remains are doubtless those mentioned byLeidy in 1896.* An examination of these shows that they belongedto a fish closely related to that known as alligator gar, usually calledLepisosteus tristoechus. However, this gar appears to the writer tobe generically distinct from the long-snouted gar and hence to becalled Atractosteus Rafinesque. The fossil materials from Willistonare referred to this genus and may be known as Atractosteus lapi-dosus. The opercular bone is made the special type of this species.It can hardly be doubted that the scales belonged to the same in-dividual fish. In case the opercular bone had the same length,proportioned to the remainder of the body, as in a specimen of Atristoechus the total length of the fish was close to 27 inches.The opercular is represented of the natural size by figure 4 of plate26. The height near the front border is 26.5 mm. ; the length is 24 mm.The corresponding dimensions of this bone in a specimen of theexisting alhgator gar are 54 mm. and 54 mm. This bone also isshown, reduced to the same size for comparison (pi. 26, fig. 5). Itbelonged to the left side of the head. The fossil bone is entire, ex-cept that a small fragment is missing, the loss of which has pro-duced the notch in the lower border. It will be seen that there aresome differences in the shape of the two bones. The greatest differ-ence is found, however, in the character of the sculpture. That ofthe fossil differs in consisting of more regular, more continuous, andmore sharply defined ridges descending from the upper angle of thebone. The ridges of the existing fish consist of a sort of networkof low ridges, especially on the front half of the bone. Also theridges of the front half are directed do\vnward or downward andbackward, while in the fossil they turn somewhat forward in theirdescent. In the existing fish the ridges of the front half are morewidely separated than those in the hinder part; in the fossil theyare narrower and more closely packed.Ten of the scales are here represented of the size of nature (pi. 28,fig. 8). It will be seen that some of them have the upper hinderborder toothed, while others have this border smooth. In theserespects they resemble the scales of the existing alligator gar (pi.28, fig. 9) except that there appear to be fewer of the teeth. BothLepisosteus osse,us and L. platystomus have the borders of all thescales smooth; at least the wiiter has not found toothed scales ineither of these species.
* Trans. Wagner Free Inst. Sci., vol. 4, page x.
112 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.50.EXPLANATION OF PLATES.Plate 26.Fig. 1. Elephas imperator. Hiadermost left lower molar. X f.2. 3. Trichechus antiquusf Lower jaw. X i-2. View of left side.3. View from above.4. Atractosteus lapidosus . Right opercular bone; outer surface. XI.5. Atractosteus tristoechus. Left opercular bone; outer surface. Reduced.Plate 27.Figs. 1. 2. Thinohadistes segnis. Left astragalus. X I-1. View of upper face.2. View of lower face.3. 4. Gnathopsis oweni. Left astragalus. X J.3. View of upper face.4. View of lower face.a, external; 6, internal, portion of tibial articular surface; c, articularsurface for navicular; d, articular surface for cuboid; e, anterior;/, posterior surface for calcaneum; g, articular surface for fibula.Plate 28.Figs. 1-3. Trucifelis fioridanus. Teeth. X 1.1. Left upper carnassial. Inner view.2. Same tooth. Outer view.3. Fragment of upper canine.4. Dinobastis series. Right canine. X 1.5-7. Felis veronis. Left upper carnassial. X 1.5. Outer view.6. Inner view.7. View of cutting border.8. Atractosteus lapidosus. Scales. X 1.9. Atractosteus tristoechus. Scales. X 1.
U. S. NATIONAL MUSEUM PROCEEDINGS. VOL. 56 PL. 26
%
Mammalian and Fish Remains from FloriFor explanation of plate see page 112
U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 56 PL. 27
r*k^ c
^- H
s
o^f
/ /; (7 g
Mammalian Remains from FloridaFor explanation of plate see rase I 12
U. S. NATIONAL MUSEUM PROCEEDINGS. VOL. 56 PL. 28
^? %
p -'
Mammalian and Fish Remains from FloridaFor explanation of plate see page 112