Front Cover
 Table of Contents
 Anatomical resume of the Haile...
 Literature cited
 Back Cover

Group Title: Fossil Testudinine turtles of Florida genera geochelone and floridemys
Title: Fossil testudinine turtles of Florida, genera Geochelone and Floridemys
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00001518/00001
 Material Information
Title: Fossil testudinine turtles of Florida, genera Geochelone and Floridemys
Series Title: Bulletin of the Florida State Museum
Physical Description: 54-97 p. : illus. ; 23 cm.
Language: English
Creator: Auffenberg, Walter
Publisher: University of Florida
Place of Publication: Gainesville
Publication Date: 1963
Subject: Turtles, Fossil -- Florida   ( lcsh )
Paleontology -- Pleistocene   ( lcsh )
Paleontology -- Florida   ( lcsh )
Genre: bibliography   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
non-fiction   ( marcgt )
Bibliography: "Literature cited:" p. 96-97.
General Note: Cover title.
 Record Information
Bibliographic ID: UF00001518
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: aleph - 000305673
oclc - 05067141
notis - ABT2273
lccn - a 63007823
 Related Items
Other version: Alternate version (PALMM)
PALMM Version

Table of Contents
    Front Cover
        Front Cover 1
        Front Cover 2
        Unnumbered ( 3 )
    Table of Contents
        Page 54
        Page 55
        Page 56
        Page 57
    Anatomical resume of the Haile series
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
    Literature cited
        Page 96
        Page 97
    Back Cover
        Back Cover 1
        Back Cover 2
Full Text
BIOLOGICAL SCIENCES Vol ume 7 N u m ber 2
Walter Auffenberg

Numbers of the BULLETIN OF THE FLORIDA STATE MUSEUM are published at irregular intervals. Volumes contain about 300 pages and are not necessarily completed in any one calendar year.
William J. Riemer, Managing Editor Oliver L. Austin, Jr., Editor
Consultant for this issue Rainer Zangerl
Communications concerning purchase or exchange of the publication and all manuscripts should be addressed to the Manning Editor of the Bulletin, Florida State Museum, Seagle Building. Gainesville, Florida.
Published 25 April 1963
Price for this issue 60e

Walter Auffenberg1
Synopsis: Recent paleontologieal work in Florida has made available a large series of fossil tortoises from one relatively small area. Two distinct species of tortoises of the genus Geochelone are represented in the known Pleistocene localities of the area, Geochelone crassiscutata (Leidy) and G. incisa (Hay). The only valid Pliocene species of this genus is Geochelone hayi (Sellards). The Florida Miocene contains only Geochelone tedwhitei (Williams).
These four species seem to represent two main evolutionary lines. Both lines are recognized as distinct subgenera, Hesperotestudo and Caudochelys; the latter is new. G. incisa is the last representative of Hesperotestudo, a line possibly beginning with G. amphithorax (Cope) of the Oligocene of Colorado, and continuing through G. osborniana (Hay) and related forms of the late Tertiary. The new subgenus Caudochelys is represented in the Florida Pleistocene by G. crassiscutata. The subgenus can be traced from G. ligonia (Cope) of the Oligocene of Colorado through G. tedwhitei of the late Tertiary of Florida, and related species of middle and western North America.
Floridemys nanus (Hay) is known only from the Florida Pliocene. Its relationships remain obscure.
1 Walter Auffenberg is Associate Director of the Biological Sciences Curriculum Study, University of Colorado. Most of his researches have dealt with the fossil history of amphibians and reptiles. Earlier contributions to the Bulletin concern fossil Terrapene and the musculature of salamanders. Manuscript submitted 26 August 1961,-Ed.
Auffenberg, Walter, 1963. Fossil testudinine turtles of Florida. Genera Geochelone and Floridemys. Bull. Florida State Mus., vol. 7, no. 2, pp. 53-97.

Introduction ....................................................... 55
Abbreviations and acknowledgments................................. 56
Material examined................................................ 56
Anatomical resume................... ............................... 58
Systematica........................................................ 69
Genus Geochelone ................................................ 69
Subgenus Caudochelys .......................................... 69
Geochelone crassiscutata....................................... 70
Geochelone hayi.............................................. 78
Geochelone tedwhitei ......................................... 80
Subgenus Hesperotestudo ........................................ 81
Geochelone incisa............................................. 82
Genus Floridemys................................................. 83
Floridemys nanus............................................. 83
Discussion ......................................................... 87
Literature cited .................................................... 96

The following forms have been described from Florida.
Miocene testudinine remains from Florida are known at present only from the Thomas Farm, Gilchrist County. Only one species is known from this locality, Geochelone tedwhitei (Williams, 1953).
Only three testudinine turtles are described from Pliocene deposits of Florida. These are Geochelone hayi, G. louisekressmani, and Floridemys nanus. It is shown later that G. louisekressmani is probably a synonym of G. hayi.
Six testudinine turtles are described from the Pleistocene of Florida. These are Geochelone incisa, G. crassiscutata, G. sellardsi, G. luciae, G. ocalana, and G. distans. G. incisa remains a valid species as described. All the others are considered synonyms of G. crassiscutata.
Another nominal testudinine genus must be considered in discussing the Florida material.
Eupachemys obtusus Leidy (1877) is based on a single peripheral of a large turtle believed to be a testudinine from the Ashley River beds of South Carolina. These beds contain fossils of various ages, but most of the terrestrial vertebrates are Pleistocene. As the fragment lacks definitive characteristics, it cannot be assigned with certainty to any particular genus. Leidy believed it represented a large aquatic emydine turtle. Hay (1908) refers it to Testudo ( = Geochelone) on the basis of size and suggests it may belong to the same species Leidy (1889) described as Testudo crassiscutata. This interpretation is also followed in the present study.
Acknowledgments and Abbreviations
The following abbreviations are used when referring to collections: ACMAmherst College Museum; AMNHAmerican Museum of Natural History; FGSFlorida Geological Survey; MCZMuseum of Comparative Zoology; UFUniversity of Florida Collections; UMMP University of Michigan Museum of Paleontology; USNMUnited States National Museum. I wish to thank the persons responsible for the collections in these institutions for the loan of^specimens in their care. Special acknowledgment is due for the criticisms and suggestions made by Claude Hibbard, Ernest E. Williams, William Milstead, Clayton Ray, William J. Riemer, and Thomas Oelrich, and to the National Science Foundation for financial assistance (G-17613).
Material Examined
This study is based on the recent discovery of much better material of Geochelone than has heretofore been available. This new material

The following forms have been described from Florida.
Miocene testudinine remains from Florida are known at present only from the Thomas Farm, Gilchrist County. Only one species is known from this locality, Geochelone tedwhitei (Williams, 1953).
Only three testudinine turtles are described from Pliocene deposits of Florida. These are Geochelone hayi, G. louisekressmani, and Floridemys nanus. It is shown later that G. louisekressmani is probably a synonym of G. hayi.
Six testudinine turtles are described from the Pleistocene of Florida. These are Geochelone incisa, G. crassiscutata, G. sellardsi, G. luciae, G. ocalana, and G. distans. G. incisa remains a valid species as described. All the others are considered synonyms of G. crassiscutata.
Another nominal testudinine genus must be considered in discussing the Florida material.
Eupachemys obtusus Leidy (1877) is based on a single peripheral of a larg turtle believed to be a testudinine from the Ashley River beds of South Carolina. These beds contain fossils of various ages, but most of the terrestrial vertebrates are Pleistocene. As the fragment lacks definitive characteristics, it cannot be assigned with certainty to any particular genus. Leidy believed it represented a large aquatic emydine turtle. Hay (1908) refers it to Testudo ( = Geochelone) on the basis of size and suggests it may belong to the same species Leidy (1889) described as Testudo crassiscutata. This interpretation is also followed in the present study.
Acknowledgments and Abbreviations
The following abbreviations are used when referring to collections: ACMAmherst College Museum; AMNHAmerican Museum of Natural History; FGSFlorida Geological Survey; MCZMuseum of Comparative Zoology; UFUniversity of Florida Collections; UMMP University of Michigan Museum of Paleontology; USNMUnited States National Museum. I wish to thank the persons responsible for the collections in these institutions for the loan of specimens in their care. Special acknowledgment is due for the criticisms and suggestions made by Claude Hibbard, Ernest E. Williarns, William Milstead, Clayton Ray, William J. Riemer, and Thomas Oelrich, and to the National Science Foundation for financial assistance (G-17613).
Material Examined
This study is based on the recent discovery of much better material of Geochelone than has heretofore been available. This new material

includes many complete but isolated elements, several plastra, carapaces, and complete shells, many associated with limb elements and vertebrae. Some specimens are so well preserved that for the first time it is possible to describe the skull and the limb and circumcloacal armor. I have examined all the type material from Florida (except that of G. louisekressmani) as well as 18 complete shells, 18 complete plastra, and numerous isolated elements.
The University of Florida Collections contains a fine series of 12 complete or almost complete shells (some with skulls) from chronologically equivalent Pleistocene localities near Haile, Alachua County, Florida (Haile VIII A, IX, and XI). The series constitutes the largest and best sample yet assembled of any Pleistocene testudinine from a single locality in North America. Also in the University of Florida Collections are complete shells from scattered localities in Florida, and a series of plastra from a locality near Reddick, Marion County, Florida (Reddick I A).
In addition to this excellent material, many scattered elements from other localities were examined.
specimens of g. crassiscutata and g. incisa examined
Type, G. crassiscutata USNM 983, along Peace Creek, near Arcadia, Hardee Co., Florida.
Type, G. incisa USNM 8821, near Ocala, Marion Co., Florida. Arredondo II, G. incisa UF uncat., two epiplastra. Eiclielberger Cave B. G. crassiscutata UF 1457, peripheral. Haile II B. G. incisa UF 3005, buckler and caudal vertebrae.
Haile VIII A. G. incisa UF 2986, 3029, 3132, partial shells; UF 3073, 3077, shell and limbs; UF 3141, complete skeleton; UF 3258, 3857, bucklers; UF uncat., two xiphiplastra; UF 3228, partial shell and limb elements. G. crassiscutata UF 3133, two scapulae; UF uncat., scapula; UF 3028, 3151, 3226, 3623, shells and limbs; UF 3139, 3623, shells; UF 3146, plastron and partial shell; UF uncat., three epiplastra; UF 3243, partial plastron; UF 3797, humerus; UF 3825, 3830, xiphiplastra; UF uncat., six xiphiplastra; UF uncat., thigh armor.
Haile IX A. G. incisa UF 3234, 3235, two shells with bucklers and parts of skeletons.
Haile XI A. G. incisa UF 3462, complete shell, buckler, and all limb elements; UF 3844, buckler.
Hillsborough River State Park. G. crassiscutata uncataloged, large shell in collection of Hillsborough River State Park.
Itchtucknee River. G. incisa UF 2713, xiphiplastron. G. crassiscutata UF 2328, xiphiplastron; UF 1994, epiplastron.
5 mi. W Jupiter Inlet. G. crassiscutata UF 3006, xiphiplastron.
Kanapaha C. G. crassiscutata UF 9907, scapula.
Mefford Cave II. G. crassiscutata UF 6523, miscellaneous parts of shell. Melbourne. G. crassiscutata, large complete shell in Amherst College Museum. Prairie Creek, G. crassiscutata UF 1816, parts of plastron.

Reddick I. G. incisa UF 4748, peripherals; UF 2545, epi- and xiphi-plastron. G. crassiscutata UF 2477, 2532, partial carapace (? crassiscutata); UF 2516, 2517, 3082, 3083, plastra; UF 2397, 2420, 2480, plastra with carapace parts; UF 2460, carapace parts; UF 2851, scapula; UF 3784, femur; UF 2475, cir-cumcloacal armor; UF 3091, xiphiplastron; UF 2598, peripherals; UF 2521, suprapygal and entoplastron.
Reddick II. G. crassiscutata UF 9908, small shell with crushed skull and limbs.
Rock Springs. G. crassiscutata USNM 11311, parts of carapace and plastron of large specimen.
Zuber. G. incisa UF 5020, epiplastron.
Because of the fine series of specimens from the Haile VIII A locality, I decided that the best approach was to study these relatively complete specimens intensively. Fragmentary specimens, isolated finds of reasonably complete individuals, and isolated single elements could then be interpreted in the light of information gleaned from this series.
Even casual observation of the shells available from this locality leads one to the conclusion that two forms are represented. Shells of both forms occur throughout the deposit, and there is no evidence of morphological intermediacy. The degrees of plastral concavity show that males and females of both forms are present. Although one form apparently attains a much greater size than the other, all small individuals can be readily separated into two groups. These data are interpreted as being evidence of differences at the species level. The type specimens of previously described Pleistocene species of this genus from Florida, in almost every case, can be placed in one group or the other.
The earliest names for identifiable fragments belonging to these groups are Geochelone incisa (Hay) and Geochelone crassiscutata (Leidy). A comprehensive description of the specimens referred to each of these species among the Haile VIII A series is given below. The reasons for allocation of the particular names used for these two forms, and the disposition of remaining named Pleistocene forms will follow. In this series five complete or almost complete specimens are referred to G. incisa, while six are referred to G. crassiscutata.
Specimens of this series placed in Geochelone incisa are much smaller than the largest placed in G. crassiscutata. See table 1. The largest specimen of G. incisa in the Haile series (UF 3141) measures 241 mm in carapacial length. The largest complete shell of G. crassiscutata from the same locality (UF 3623) measures 1210 mm in carapacial

Table 1. Shell measurements (in millimeters)
G. incisa G. crassiscutata
3462 3141 3077 3073 3029 2986 3235 3151 2480 2397 3139 3226 2420 3028
Carapacial length 241 231 228 224 216 212 327 192 180 168
Plastral length 264 205 214 204 206 192 196 304 259 257 180 169 164 153
Entoplastral length 52 41 42 40.5 41 39 33 72 59 55 46 38 34 39
Entoplastral width 46 36 36 36 40 40 34 71 57 55 39 39 31 32
Cular length 46 34 36 37 30 30 34 54 39 43 31 27 24 28
Humeral length 37 36 48 32 35 34 33 54 51 45 31 30 29 29
Pectoral length 11 18 12 10 7.5 12 11 19 20 14 9 7 12 13
Abdominal length 94 66 59 74 65 70 68 115 79 89 68 63 59 52
Femoral length 42 38 32 32 28 32 28 41 43 31 23 21 19 24
Anal length 16 10 9 8 12 7 12 17 22 21 9 13 6 19 12
Xiphiplastral notch depth 28 22 23 21 23- 12.5 20 24 12 13 14 12 8
Xiphiplastra! external height 28 23 22 21 29 19 18 24 21 21 16 14 13 12
Xiphiplastral length 33 35 35 30 35 29 32 51. 43 45 33 30 33 26
Epiplastral length 51 45 50 41 49 39 40 68 60 51 29 32 32 31
Epiplastral width 47 44 43 49 40 39 39 63 57 49 38 34 31 33
Epiplastral lip thickness 36 29 28 27 31 25 21 35 31 31 17 22 19 21
Epiplastral lip length 36 33 40 31 37 39 45 36 36 27 23 19 21
Nnchal scute length 20 17 15 14.5 11 10 11 20 16 11 10 10 9
Nuchal scute width 3 8 6 7 3 9 4.5 13 9 9 22
Pygal length 30 29 31 22 27 22 44 21 19 20
Pygal dorsal width 40 39 46 31 40 28 48 26 26
Pygal ventral width 27 23 23 19 24 19 36 19 15 14 24
Second suprapygal length 24 19 21 19 31 20 27 15 14
Second suprapygal width 62 46 55 41 57 44 45 24 22 34

length, though fragments of much larger specimens have been found. Specimens referred to in G. incisa are more rugose than similar-sized or even larger specimens of G. crassiscutata. Sulcal growth patterns are much more evident in the former, the central areas of the costals are more raised, and small bosses are developed in several parts of the shell, particularly on or near the peripherals. The latter are considerably thickened, particularly the posterior members.
The plastron is also quite different in the two species. In G. incisa the xiphiplastral notch is much deeper (fig. 1), with the free tips more pointed. The epiplastral beak is proportionately thicker. In
1 13 i Reddick I
I Mailt 3znea
8|_( Haile 2dia
AGE 1.0 15 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7C 7.5 8.0 8.5
median length xiphiploslron depth of xiphiplastral notch
Figure 1. Specimens of Geochelone from the Pleistocene of Florida can be separated on the basis of the proportional depth of the xiphiplastral notch. The Great Plains species, G. turgida and G. rexroadensis, seem most closely related to G. incisa and G. crassiscutata respectively.
most specimens the tips of the beak are slightly divergent, and a keel may be present on each of the two elements, both dorsally and ven-trally, and running out to the tips.*The lateral parts of the hyo-, hypo-, and xiphi-plastral elements are thickened much more in G. incisa than in G. crassiscutata (fig. 2).
pectoral girdle and limbs
The scapulae of G. crassiscutata and G. incisa are similar. In G. incisa the procorocoid process is proportionally shorter than in G. crassiscutata. The scapular portion of the glenoid cavity is proportionally

wider in G. incisa than in G. crassiscutata (fig. 3). The included angle of the blade of the corcoid in G. incisa is usually smaller than in G. crassiscutata (measurement made with the apex of the angle located
4 N 6 ? h a i y H < [ is U i 'a * vi] D H 'M I 2 S3 = F>j crassiscutata Jj incisa a. "a rO 5 rexroadensis turgida
Median length xiphiplastTon External hetghl xiphiplaslrori
Figure 2. G. incisa and G. crassiscutata are readily separated on the basis of the proportional external height of the xiphiplastron.
on the shaft). Holman (1959) suggests that two humeral forms in the Florida Pleistocene indicate the presence of two species of Geochelone. However, information obtained from complete specimens of all sizes of the two species indicates that the differences are usually subtle, and much ontogenetic and individual variation exists. The tuberosity in G. incisa seems higher than in G. crassiscutata in the few individuals, of equal size which can be compared. The head is more rounded and set at less of an angle in G. crassiscutata. In early members of the Incisa group (G. oshorniana, G. orthopygia), the humeral head is like that in G. crassiscutata.
Lack of material prevents comparison of the ulna and radius.
pelvic girdle
The lateral extent of the ventroposterior process of the ischium is greater in G. incisa than in G. crassiscutata. The extension is more acute and more twisted in the former. The anteroexternal process of the pubis tends to have a shorter articular surface in G. incisa, though there is age variation. Small specimens have a proportionally longer

surface. The curvature between this process and the midline is greatest in G. incisa (fig. 3).
Fhsujhk 3.
Top left) Six fust:d and expanded caudal vertebrae of a specimen of G. incisa, (UF 3077).
Top right) Proportional width of the scapular portion of the glenoid cavity in (left) G. incisa { UF 3077) and (right) G. crassiscutata (UF 3151).

The distal end of the ilium is generally thinner and less twisted in end view in G. incisa than in G. crassiscutata (fig. 3). Small specimens of the latter have ilia that are less twisted, but as far as is known, they arc never as straight as in G, incisa.
The femur of G. incisa is more curved from the side, and the epi-condyles are developed more than in specimens of G. crassiscutata of similar size. The intertrochanteric iossa is relatively much larger in G. incisa. The angle between iho head and the external surface of the greater trochanter is greater in G. incisa than in G. vrassicutata.
In the largest available specimen of G. incisa (UF 3462), the fibu-lare is strongly co-ossified with the astragalo-calcancum. A slightly smaller specimen of the same species (UK 3141) has this co-ossification on the left limb, but on the right the bones are Oflly weakly united. In a sirnilaT-sized specimen (UF 3073} they arc not fused. In a single specimen of G, crassiscutata in which the foot bones are preserved (UF 3.151 ) these bones are not fused, Whether such fusion occurs in larger specimens of G. crassiscutata h unknown. Young specimens of G. incisa prohably lack this fusion.
In UF 3462 (G. incisa) tarsal 3 is missing. Tarsals 4 and 5 seem ir> be fused. Such fusion is not present in the best hindfoot of G, crassiscutata examined (UF 3151). Metatarsal 3 and phalange III are fused m UF 3462.
In G. incisa the terminal portion of the tail is expanded to support the heavy, co-ossified supracaudal ossicles (fig. 3). The last 5 vertebrae of G. oshorniamt have lengthened lateral processes. In G. ortho-pygia at least 6 vertebrae show these elongated processes. The last 8 vertebrae are so modified in the specimens of G. incisa examined. Thus, the available material of this phylotic line suggests an increase from the Miocene to Pleistocene in the number of vertebrae supporting the supra caudal buckler.
The inference by Hay (1908:426) that the lengthened transverse processes of G. osborniana differ basically from the "distinct, true caudal ribs" of Gopher us polyphemus is apparently in error. In UF 3462, an excellent specimen of G. incisa, most of the processes are
Middle left) Shape of the distal end of the ilium in (left} G. incisa (UF 3077)
and (right) G. crassiscutata (UF 3151 ). Middle right) The supracaudal buckler of G. incisa (UF 3077). Lower left) The forelimb armor of G. incisa (UF 3141).
Lower right) Pelvic girdle of (left) G. crassiscutata (UF 3151) and (right) G. incisa {UF 3141).

firmly ankylosed to the vertebrae. However, several of the vertebrae clearly show a thickened seam at the juncture of the base of the process and the vertebral body. Furthermore, a few of the vertebrae anterior to the buckler lack the processes, whereas adjacent members possesses them. Where they arc missing the articular surface of the caudal rib is clearly evident. In a small specimen of G. incisa (UF 3141) all the vertebrae anterior to the buckler lack the ribs but have the articular surface. Tn the vertebrae supporting the buckler the ribs arc all firmly fused to the centra. The ribs are not fused to the centra in specimens of G. crassiscutata of the same size. However, in large specimens of this species these ribs are all firmly fused to the vertebrae. Thus the character seems to be associated with size, at least in the more anterior vertebrae,
Complete cervical series of both G. incisa and G. crassiscutata are available. Most individual members of the series show no differences between the two species, but the second vertebra does show a few differential characters. In G. incisa the prezygapophysial articular surfaces are almost horizontal when viewed from the front. In G. crassiscutata the surfaces are at an angle1 of about 45 degrees, In G. incisa the anterior articular surfaces of this element are sometimes biconcave (2 out of 4 complete series associated with complete shells). In G. crassiscutata the surfaces are biconvex in all 4 specimens examined. Manv more vertebral scries associated with shells are needed
Fh;l-kg 4, The skull of G. incisa ( UF 3M1 )

to validate this suspected proportional difference in shape of the articular surfaces,
A complete skull of G. incisa, associated with the rest of the skeleton and shell (UF 3141) from Haile VIII A, is rather long and narrow (fig. 4, and table 2), but less so than that in G. osborniana. The sides are almost parallel lrom the squamosal processes to the posterior border of the orbits. Near the middle of the orbits the maxil-laries converge to the snout and are slightly concave from above. From the side the dorsal surface is slightly concave behind the orbits, and strongly convex above? and in front of them. The cutting edge of the maxilla is concave and finely serrated. There is a low, wide notch with slightly larger serrations on either side anteriorly. The squamosal processes are slightly compressed laterally, less so internally.
Tabj.r 2. Measurements of the skull of G. incisa (UF 3141)
Snout to end of supraoecipital process 51 JIlTIl
Snout to occipital condyle 42
Width at base of quadrates 35
Width at posLerior end of tria_\iliac 33
Width of iritcrorbit&l spaee 14
Width of nasal opening 12
Anteroposterior diameter of orbit 14
Maxillary ed^e to highest part of front;tl 18
Median length of prefrontals 9
Median length of irontals 11
Width of jugal art:h 7
The upper outer border rises above the upper surface of the par occipital.
The height of the skull from the cutting edge of the maxilla to the upper surface of the frontals is contained in the length from the snout to the condyle 2.3 times. The anteroposterior extent of the otie region is 12 mm, and is contained in the length of the skull to the condyles 3.5 times.
The palatal region is highly vaulted, the median fossa having a, width of 13 mm. The narrowest part of the pterygoids is II mm. The alveolar surface of the maxilla is provided with 2 longitudinal ridges and 2 grooves. The outer, larger ridge is separated from the cutting edge by the wider groove. The ridge is sharp, slightly serrated, and runs forward and inward to the premaxillary, is then directed anteriorly to the cutting edge of the same element. The inner ridge is higher than that in G. osborniana. It is broader posteriorly

Figuhe 5. The circumcloacal armor of G. crassiscutata (UF 2475).
than anteriorly and runs along the inner edge of the maxillary. There is no median premaxillary ridge.
The skull of G. incisa is more like that of G. osborniana than that of G. gilberti, but the jugal arch is wider still than in the former. The otic region, like that in G. osborniana, is less elongated than in G. gilberti. The skull of G. impensa differs from G. incisa in the structure of the maxillary ridges, in having a narrower jugal arch, a narrower palatal fossa, wider prefrontals, and longer squamosal processes. The skull of G. orthopygia is like that of G. osborniana, and thus similar to that of G. incisa. In G. orthopygia the skull is wider than in G. incisa; the prefrontals are wider, and the frontals proportionally shorter. The cutting surface of the mandible is longer in G. incisa than in G. orthopygia.
A small specimen of G. crassiscutata in the University of Florida Collections from Reddick II (UF 9908) includes a smashed skull. Little can be determined from this specimen except that the frontal is proportionally longer than in G. incisa.
limb armor
UF 3151, a half-grown G. crassiscutata shows some of the armor of both the front and hind limbs. The bottom of the hind foot is covered with fairly large, flat, bony plates. Those at the posterior periphery of the foot are much larger, and more spurlike than those over the rest

Figure 6. A diagramatic illustration of the posterior limb and caudal armor shown in black in (top) G. incisa and (bottom) G. crassiscutata.
of the foot. The armor of the front limbs below the humerus is unknown in G. crassiscutata. However, a number of large, lanceolate, flattened scutes found with several large specimens suggests that a part of the forelimb was armored.
The thigh armor of G. crassiscutata is now well known. Two groups of ossicles from this region, associated with shells, are available. Both specimens are large. In each, several ossicles are fused together to form a plate on the posterior surface of the thigh. The plate is composed of approximately 12 large, highly peaked ossicles (figs. 5, 6). Around this plate the ossicles are usually smaller, though a few large spurs occur between the plate and the base of the tail.
The heel armor of G. incisa is unknown. The armor of the forelimb shows in UF 3141 (fig. 3) and seems to be composed of proportionally larger ossicles than in G. orthopygia. In fact, the ossicles on the outer edge of the limb are proportionally larger than in most extant testu-dinines except for a few Old World forms. The largest ossicle is contained in the length of the ulna 2.5 times.
As far as known the ossicles on the posterior surface of the thigh of G. incisa are not fused as they are in G. crassiscutata. Instead, the area seems to be covered with large, loose, conelike ossicles. As in large specimens of G. crassiscutata, spurs are present between this area and the base of the tail.

caudal buckler
One of the most diagnostic features of G. incisa is a fused series of dermal scutes immediately above the tail (fig. 3). The entire series comprises a subcircular patch composed of from 30 to 40 scutes. All the elements are quite firmly ankylosed in larger specimens. The scutes are pentagonal, hexagonal, or octagonal. Although there is no definite alignment, the central scutes are usually largest. Each scute near the center usually has a small spur, frequently off center. The scutes close to the edge are more conelike.
On the basis of several complete specimens in which this buckler was found in place, in life it was obviously located near the median posterior edge of the carapace. The pygal and the adjacent carapacial bones possess a transverse depression on their inner surfaces. It is presumed that the plate was hinged to the underside of the carapace along its upper edge (fig. 7).
Figure 7. Cross-sectional views of the posterior parts of the shells of (A) G. crassiscutata and (B) G. incisa showing the diagnostic features of the caudal area.
The caudal vertebrae are closely appressed to the underside of the buckler, where the compressed and elongate transverse processes sometimes fit into shallow grooves.
A supracaudal buckler of fused scutes, with similarly modified vertebrae has been described in G. orthopygia and G. osborniana (Hay, 1908), and in G. turgida (Oelrich, 1957), as well as in some less well known forms. In all these species the unit is composed of fewer and more simple elements.
The tail of G. incisa was obviously short and probably did not show externally in life (fig. 8). The last few caudal vertebrae are much

Figure 8. Presumed appearance, in life, of the tail of (A) G. crassiscutata and (B) G. incisa.
compressed caudocranially and do not extend beyond the posterior or ventral edge of the buckler.
Evidently two groups of tortoises of the genus Geochelone were present in T?\oxida during the Pleistocene. "Each group is represented by a single species, G. incisa and G. crassiscutata. These groups seem to correspond to the two lines recognized in western North America G. orthopygia-G. osborniana and G. Ugonia-G. rexroadensis. These two groups are so distinct that it seems best to recognize them as two separate subgenera. G. osborniana (Hay) (figs. 27, 28) is the type species of the subgenus Hesperotestudo, including G. orthopygia (fig. 29), G. turgida (fig. 9), G. riggsi (fig. 31), G. incisa (figs. 22, 23, 24), and all of their relatives. G. uintensis, G. tedwhitei (fig. 21), G. rexroadensis, G. crassiscutata (figs. 10, 11, 12, 13), and all of their relatives seem to form a natural group here designated as a new subgenus.
SYSTEM ATICS Genus Geochelone Caudochelys, new subgenus
Type species. Geochelone crassiscutata (Leidy)
Definition. A Nearctic subdivision of the genus Geochelone with a narrow nuchal scute, entoplastron about as wide as long, pectoral scutes reduced along the midline, and limbs and tail heavily armored with dermal ossicles. Ossicles above the tail are never fused to form a supracaudal buckler, and the vertebrae are normal, not

Figure 9. A series of plastra of G. turgida, showing the projected and keeled epiplastral beak also found in G. incisa.
compressed or fused, without greatly elongated transverse processes. Eocene to Pleistocene of North America.
Geochelone (Caudochelys) crassiscutata Figures 10,11, 12,13
Testudo crassiscutata Leidy, 1889.
Testudo sellardsi Hay, 1916. Loomis, 1927; Williams, 1950.
Testudo luciae Hay, 1916. Loomis, 1927.
Testudo ocalana Hay, 1916.
Testudo distans Hay, 1916.
Gopherus ocalana (Hay) Williams, 1950.
Geochelone sellardsi (Hay) Ray, 1957.
Geochelone luciae (Hay) Ray, 1957.
Geochelone ocalana (Hay) Ray, 1957.
Type. USNM 983, the greater portion of the right epiplastron, a part of the right hyoplastron, parts of the posterior lobe of the plastron, the left femur and the left tibia (fig. 10).
Type locality and horizon. Peace Creek, near Arcadia, Hardee County, Florida; Pleistocene, Recent, or both, but probably only the former.

Figure 10. Type of Geochelone crassiscutata (USNM 983), (A) right epiplas-tron, (B) xiphiplastron.
Figure 11. Subadult G. crassiscutata (UF 3151) from Haile VIII A, Alachua County, Florida

Figure 13. Plastra of subadult specimens of G. crassiscutata from Reddick I A, Marion County, Florida. (A) UF 2517, (B) UF 2480.

Diagnosis. A species of giant Pleistocene testudinine, differing from G, incisa in being larger, having a thinner, less rugose shell, a shallower xiphiplastral notch, a less well developed and simpler epiplastral beak, a longer tail, and, in large specimens, a small fused patch of dermal ossicles on the posterior surface of the thigh. It also differs in lacking a caudal buckler.
The major differences between the two species of Florida Pleistocene tortoises are:
1) 2)
3) 4)
6) 7)
8) 9)
G. incisa
Carapacial length to 231 mm Supracaudal ossicles fused into a rounded caudal buckler (figs. 3,6)
Caudal vertebrae fused, compressed anteroposteriorly, transverse processes attenuate (fig. 3)
Shell, and particularly the peripherals, thicker proportionally, rugose (particularly in males) Xiphiplastral notch deep, included angle less obtuse (fig. 1)
Epiplastral beak thicker, frequently strongly keeled Xiphiplastron proportionally higher at outer edge of hypohyoplastral suture (fig. 2) Width of first suprapygal greater (fig. 14) Entoplastron narrower (fig. 15)
G. crassiscutata
Carapacial length to at least 1150 mm Supracaudal ossicles not fused into a caudal buckler (figs. 5, 6)
Caudal vertebrae not fused, not noticeably compressed, transverse processes not as attenuate
Shell thinner proportionally, smoother
Xiphiplastral notch less deep, included angle more obtuse (fig. 1)
Epiplastral beak smaller, thinner, rarely keeled, and then only slightly Xiphiplastron proportionally lower at outer edge of hypohyoplastral suture ( fig. 2)
Width of first suprapygal less (fig. 14)
Entoplastron wider (fig. 15)
- 1 1 1 II 1 1 1
incisa crassiscutata
1 I i 1 II I 1
Median length plastron Width first suprapygal
Figure 14. Proportional width of the first suprapygal in G. incisa, G. crassiscutata, and G. turgida.

2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
Plastral length Width entoplastron
Figure 15. Proportional width of the entoplastron in G. incisa, G. crassiscutata, and G. turgida.
Both species exhibit considerable variation in many of these characteristics, usually associated with sex or size. However, some evidence in G. crassiscutata (in which small series are available from several localities) suggests that minor chronological or geographical differences occur in certain shell ratios (fig. 16). The two available specimens of G. incisa from Haile X, Alachua County, have caudal bucklers with ossicles considerably more flattened than in any specimen from Haile VIII A, Alachua County. If valid, such differences are probably indicative of chronological rather than geographical differentia-
ReddickIA -
Haile3ZDTA- o
35 40 45 50
Length of epiplastron (in mm)
Figure 16. An example of minor differences in shell ratios in two populations of G. crassiscutata.

tion, as the localities are close to one another. Until larger series are available from additional localities such presumably minor differences cannot be critically evaluated.
Geochelone crassiscutata (Leidy, 1889) is described on the basis of fragments of three individuals collected from the shoals of Peace Creek. The original collection of specimens made by J. Francis Le-Baron was numbered 983 to 986 in the United States National Museum. As Hay (1908) restricted the type material to that figured by Leidy, the type number should be USNM 983 (the number of the figured xipiplastron, etc.), rather than USNM 985, as stated by Hay. Most of the nontype material of the same lot is catalogued as USNM 6730.
The type material includes a large part of the right epiplastron (fig. 10), a part of the right hypoplastron, parts of the posterior lobe of the plastron, the left femur, and the left tibia. These remains suggest the complete turtle probably had a carapacial measurement of approximately 1520 mm. Hay referred to the same species a number of other fragments of large testudinine turtles in the United States National Museum and the American Museum of Natural History. The preserved elements, particularly the xiphiplastron, suggest that this species represents the second major phyletic line of Geochelone in the Pleistocene of Florida, being closest to the Miocene G. tedwhitei (Williams).
Figure 17. Type of Geochelone sel-lardsi (USNM 8817), a left xiphiplastron.
Geochelone sellardsi (Hay, 1916), is based on part of a xiphiplastron and parts of a carapace of a large testudinine turtle (fig. 17) collected at Vero, Indian River County, Florida (USNM 8817, formerly FGS V-1831). The diagnosis indicates that G. sellardsi is similar to G.

crassiscutata (Leidy), but differs from it (Hay, 1908) in having the outer face of the anterior part of the thickened xiphiplastral border flat or concave instead of convex. The thickness of the anterior end of the border is contained in the distance to the bottom of the xiphiplastral notch 3.6 times, instead of 3 times as in G. crassiscutata. Other characters mentioned by Hay as important in separating the two species are the extent of horn covering the upper surface of the xiphiplastral lobes, and the sculpturing of the lower surface of the lobes. Hay thought that the degree of sculpturing might vary between individuals. An almost complete specimen presumably representing this species was described by Loomis (1927).
The characters chosen as diagnostic of the species are now all known to be variable within a sample of one species from one locality. The name G. sellardsi is here referred to the synonymy of Geochelone crassiscutata Leidy.
Geochelone luciae (Hay, 1916) has as the type a part of the right hypoplastral element (USNM 8818, formerly FGS V-1807) of a large tortoise from Vero, Indian River County, Florida. Hay states that this species possibly attained the size of G. crassiscutata. He says it differs from the latter in having a thinner wall along the border of the base of the posterior lobe. However, on the type the articular surface between the hypoplastron and the xiphiplastron is broken away. Therefore, Hay's comparison of the shape of this area in G. luciae and G. sellardsi is not warranted. Furthermore, the element is one
Figure 18. Type of Geochelone ocalana (USNM 8822), a right epiplastron, (A) dorsal, (B) ventral.

which shows considerable variation in thickness. Evidence shows that this variation is both individual and ontogenetic within a single species from a single locality. As no feature clearly separates G. luciae from G. crassiscutata, the name is here placed in the synonymy of the latter.
Geochelone ocalana (Hay, 1916) is the right half of an epiplastron (fig. 18) collected from a fissure in a quarry of the Florida Lime Company, Ocala, Marion County, Florida. The type is USNM 8822 (formerly FGS V-4299). According to Hay, the available material places this species closest to G. crassiscutata but differs from it in having a thicker epiplastral beak. The thickness of this element is extremely variable and almost useless in identification of Geochelone remains from Florida. A hypoplastral and first pleural from the same locality were provisionally placed in this species by Hay. Williams (1950) incorrectly placed the species in the genus Gopherus, but indicated that the material was so fragmentary that positive identification was impossible. The present study indicates that G. ocalana is best placed in the synonymy of G. crassiscutata.
Figure 19. Type of Geochelone distans (USNM 8819), an entoplastron.
Geochelone distans (Hay, 1916). A complete entoplastron (USNM 8819, formerly FGS V-4289) is the type (fig. 19). It was collected at pit 5 of the Florida Lime Company at Ocala, Marion County, Florida. The distinguishing feature of the element is that the pectoral scutes extend anteriorly onto the entoplastron, whereas in nearly all other species these scutes have their anterior border just behind this element. The xiphiplastron is clearly from a large specimen, presumably possessing a thinner plastron than that in the type of G. crassiscutata. Hay was unable to compare the element with that oL-any other Pleis-

tocene species from Florida, except G. ocalana, from which it differs considerably.
The species G. distans is here tentatively placed in the synonymy of G. crassiscutata on the basis that the presumed diagnostic character occurs as a variant in several species of tortoises.
Geochelone (? Caudochelys) hayi
Testudo hayi Sellards, 1916.
Testudo louisekressmani Wark, 1929.
Gopherus hayi (Sellards) Williams, 1950. Ray, 1957.
Type. USNM 8815 (formerly FGS V-5001) a large portion of the posterior parts of both the carapace and plastron.
Type locality and horizon. Amalgamated Phosphate Company pit, Brewster County, Florida; Pliocene, Bone Valley Gravel formation.
Diagnosis. A large species of Pliocene testudinine turtle differing from the dwarf Floridemys nanus in being much larger, and in having the gular sulcus at an angle with the midline of the plastron, rather than perpendicular to it. The xiphiplastral notch is deeper and more acute than in G. crassiscutata. The exterior wall of the hypoplastron of G. hayi is vertical, while in G. crassiscutata it slopes inward. The carapace of the type of G. hayi, though larger than that of G. crassiscutata, is much thinner.
Geochelone hayi represents a fairly large testudinine, estimated to have been about 1500 mm in length. The second neural is octagonal,
Figure 20. Rear portion of the carapace of the type of Geochelone hayi _(USNM 8815) showing the correct shape of the second suprapygal.

and neurals 4 to 8 are hexagonal. The proximal end of the second pleural is slightly reduced in width and is in contact with the second neural only, while the third pleural touches neurals 2, 3, and 4. The first suprapygal is large and rests on the eleventh marginal and the pygal. The second suprapygal is considerably reduced (Sellards's [1916] figure of the type showing the posterior portion of the carapace is in error, as the second suprapygal is wider than originally illustrated [fig. 20]).
The species G. hayi was recently placed in the genus Gopherus (Williams, 1952). This interpretation is probably incorrect and G. hayi is here referred to the genus Geochelone, and tentatively to the subgenus Caudochelys.
Geochelone louisekressmani (Wark, 1929). This species is described on the basis of a disarticulated partial shell, including parts of the right and left halves of the epiplastron, fragments of the hypoplas-tron, a humerus, a pygal, several fragments of the xiphiplastron, marginals, pleurals, and vertebral scutes. All this material was collected from pit 5 of the Amalgamated Phosphate Mining Company near Brewster, Florida. Stratigraphic data are lacking, though the specimen is thought to have come from the Bone Valley Gravel formation.
Many opinions on the age and manner of deposition of this formation are published. This apparently is because of the complexity of the lithologic units. Most workers now agree that both Miocene and Pliocene vertebrates are represented in its fauna; almost all the marine forms are considered Miocene, and the terrestrial species Pliocene. Of particular importance is the fact that the same mine has also produced many remains of Mastodon and Chlamytherium (Wark, 1929), both Pleistocene genera. Local deposits of Pleistocene age apparently occur throughout this entire region; normally as shallow lenses near the top of the section and probably representing stream or pond deposits. G. louisekressmani conceivably could have originated from these superficial beds. It is here placed in the Pliocene, but with reservations.
The type is a poor reconstruction of a poor specimen. Wark states that the carapace is represented by the "right forepart." Unfortunately the type could not be located, but examination of the published figure suggests that it is the rear part, embracing parts of both the right and left sides. The reconstructed sulci in Wark's figure are completely in error.
The type was collected at the same mine and at the same time as

the material described by Sellards as Testudo hayi. Sellards refers several other fragments collected from this mine to T. hayi. Wark makes no reference to the type or referred material of T. hayi. The original number of the type of G. louisekressmani and also that of T. hayi was FGS V-5001. Hence Wark's material was either part of the type material of T. hayi, or at least material referred to T. hayi by Sellards. Unfortunately, Sellards does not describe this additional material adequately enough to be positive.
The only diagnostic characters mentioned by Wark are great size, and lack of symmetry of the peripherals, pleurals, and neurals. The type material of G. hayi does not represent a small testudinine, though one considerably smaller than the type of G. louisekressmani Because nothing is known of the maximum size attained by G. hayi, the character is worthless. The lack of symmetry in various elements of G. louisekressmani can be explained by Wark's misinterpretation of the part of the shell represented in the type material. In proper position these elements not only seem quite symmetrical, as far as can be determined, but are similar to the same elements in the type of G. hayi. Therefore, the name Testudo louisekressmani Wark is here referred to the synonymy of Geochelone hayi (Sellards).
The University of Florida Collections has additional material referred to this species from several localities. This includes several isolated elements and fragments of the shell of an individual collected at Haile VI A, near Haile, Alachua County, Florida; a few peripherals from the Camp 12 Mine, near Holder, Cijtrus County, Florida; and a few fragments from the McGehee property, a few miles north of Newberry, Alachua County, Florida. None of these fragments gives any additional information on variability in the specific characters of G. hayi.
Geochelone (Caudochelys) tedwhitei Figure 21
Testudo tedwhitei Williams, 1953. Geochelone tedwhitei (Williams) Ray, 1957.
Type. MCZ 2020, a complete plastron.
Type locality and hoeizon. Thomas Farm, Gilchrist County, Florida; Arikareean, Lower Miocene; Hawthorne formation.
Diagnosis. A medium-sized Miocene species of testudinine with gulars more triangular and the ends smaller than in G. crassiscutata. It differs from G. hayi in being smaller and in the shape of the supra-pygal.

Remarks. Most of the following data comes from Williams (1953). As far as known, G. tedwhitei does not exceed 400 mm in plastral length. The length of the median sulcus of the abdominals is 6 to 7
Figure 21. Geochelone tedwhitei. (A) Carapace. (B) Type plastron (MCZ 2020). Both from Lower Miocene, Thomas Farm, Gilchrist County, Florida.
times that of the pectoral scutes. The gulars are more nearly triangular and the anals smaller than G. crassiscutata. The well developed nuchal scute reaches the anterior margin, and the costovertebral sulci are not deeply incised. The free margins are less reverted than in G. crassiscutata. Several plastra and an almost complete carapace are available. G. tedwhitei is apparently most closely related to G. farri and G. ducatelli.
Subgenus Hesperotestudo Williams
Type species. Geochelone osborniana (Hay).
Definition. A Nearctic subdivision of the genus Geochelone with a narrow nuchal scute, an entoplastron about as wide as long, reduced pectoral scutes, and limbs and tail heavily armored with dermal ossicles. Those above the tail are fused to form a supracaudal buckler; the tail vertebrae are compressed, sometimes fused. The transverse processes of the caudal vertebrae are greatly elongated. (?)Eocene to Pleistocene of North America, and "Tertiary" of Asia.

Figure 22. Type of Geochelone incisa (USNM 8821), a left xiphiplastron.
Diagnosis. A species of Pleistocene testudinine differing from G. crassiscutata in being smaller, more rugose, with a thicker shell, sometimes with bosses on some of the peripherals and on the bridge; the epiplastral beak is proportionately thicker, frequently with, a keel above and below, and with acute tips that are more divergent than in G. crassiscutata; the xiphiplastral notch is deeper, the enclosed angle being less than in G. crassiscutata; the entoplastron is proportionally longer in G. incisa; the interanal sulcus is shorter; the buttress at the outer anterior edge of the xiphiplastron is proportionally much higher. In life the tail may not have been noticeable externally; the last few caudal vertebrae are shortened and fused; the transverse processes are larger than in G. crassiscutata. Ossicles of the posterior surface of the thigh are not fused into a single unit, and a number of ossicles above the tail are fused into a plate that is absent in G. crassiscutata.
Geochelone (Hesperotestudo) incisa Figures 22, 23,24
Testudo incisa Hay, 1916.
Gopherus incisa (Hay) Williams, 1950.
Geochelone incisa (Hay) Ray, 1957.
Type. USNM 8821, a complete left xiphiplastron (fig. 22). Type locality and horizon. Pit 5, Florida Lime Company, Marion County, Florida; Pleistocene.

Figure 23. Shell of adult G. incisa (UF 3029) from Haile VIII A, Alachua County, Florida. (A) lateral, (B) posterior, (C) ventral, (D) anterior.
The type is from the same locality as the type of G. ocalana. The diagnosis emphasizes the thick, heavy nature of the xiphiplastron. It is provided with a deep notch posteriorly. The terminal processes are acute, and the anal scute is very short at the midline.
This species is distinct and valid, and can now be completely diagnosed. A number of complete shells, some with limbs and even limb armor, are available from several localities in Florida. It obviously belongs to the Osborniana-Orthopygia-Turgida line, and represents the previously missing Pleistocene representative postulated by Williams in 1953.
Genus Floridemys Floridemys nanus
Figure 25
Bystra nana Hay, 1916.
Floridemys nanus (Hay) Williams, 1950. Williams, 1952; Loveridge and Williams, 1957.
Type. USNM 10247, a complete shell of an adult specimen of a distinctive dwarf species of Pliocene testudinine, crushed on the left side of the carapace, and pushed dorsally so that in side view it looks

considerably higher than it probably was in life. Originally the type was in the collection of Dr. Henry G. Bystra of Brooksville, Florida.
Type locality and horizon. Collected from a hard-rock phosphate mine, near Holder, Citrus County, Florida; Pliocene (or possibly Miocene), Alachua formation.
Diagnosis. A small species of tortoise distinguished from all other described forms by a transverse gulohumeral sulcus. Other important characters are the truncated epiplastral beak and the straight humero-abdominal sulcus that is directed posteriorly from its median origin.
Description. The type of Floridemys nanus has been illustrated by Hay (1916), though the figure leaves something to be desired for tracing articulations and sulcal patterns. An examination of the specimen under different types of light, and both wet and dry, disclosed a number of characters not determinable from Hay's illustration, and not mentioned in the description. The specimen is therefore redescribed.
A dwarf species of testudinine turtle known from a single shell of an adult female turtle 105 mm carapace length. The sulcal growth patterns are moderately developed. The pleurals are alternately wide and narrow at their distal ends. The first pleural is in contact with only the first neural, which is subrectangular in shape. The second neural is hexagonal with the short lateral sides directed posteriorly; the third
Figure 25. Type shell of Floridemys nanus (USNM 10247) slightly reconstructed (the original is crushed laterally). (A) lateral, (B) anterior, (C) ventral, (D) posterior.

is apparently quadrangular; the fourth presumably octagonal; and the remaining ones hexagonal with the short sides directed anteriorly. The first suprapygal is bifurcate, the second subtriangular. The pygal is wider dorsally than ventrally. The epiplastral lip is truncated, and does not project beyond the carapace. The gular sulcus does not reach the entoplastron and is almost perpendicular to the main axis of the plastron. The epiplastral bones are short and wide, compared to most species of tortoises. The entoplastron is slightly longer than wide. The gular scute is much shorter than the humeral, but slightly longer than the pectoral. The humeral is also longer than the femoral, and the anal is about as long as the pectoral. The xiphiplastral notch is wide and deep, almost as deep as the length of the xiphiplastral bone along the midline. The hyoplastron is much shorter along the midline than the hypoplastron.
Table 3 gives the pertinent measurements of the type specimen. Some measurements differ slightly from those given by Hay (1916), owing to differences in the way measurements were taken, or to better subsequent preparation of the specimen.
Table 3. Shell measurements of the type of Floridemys nanus
Carapace length____________________ 105 mm
Carapace width (estimated) 80
Carapace height (estimated) 53
Plastron length (midline) 96
Xiphiplastral notch depth ____ 12
Xiphiplastral notch width____ 26
Xiphiplastral length______________ 15
Hypoplastral length______________ 32
Hyoplastral length ______________ 21
Entoplastral length ______________ 18
Entoplastral width ______________ 17
Epiplastral length ________________ 16
Epiplastral width ........____ 17
Anterior lobe base width ____ 44
Posterior lobe base width .... 55 Lip, width across anterior
end ____________________________________ 23
Lip, greatest thickness__________ 11
Lip, greatest length______________ 16
Xiphiplastral external height
at hypoplastral articulation 11
Xiphiplastral greatest width 24
Gular length ________________________ 8
Humeral length _____________________ 20
Pectoral length______________________ 6
Abdominal length ________________ 39
Femoral length ____________________ 15
Anal length __________________________ 8
Pleural 1 distal length__________ 15
Pleural 1 proximal length .... 12
Pleural 2 distal length__________ 15
Pleural 2 proximal length____ 6
Bridge length ______________________ 54
Floridemys nanus is based on a small shell manifesting a single aberrant character. No additional material of this interesting dwarf turtle is available. Williams (1950) is tempted to regard this specimen as an individual variant and, in the absence of other skeletal remains, is apparently not even sure of its testudinine affinities.

Despite the temptation to regard this specimen as an individual variant, the name must be provisionally retained. The aberrant character, the transverse gular-humeral sulcus, is in the form exhibited here, rare in tortoises. .. When so rare a character is coupled with extremely small size ... it becomes necessary to withhold judgment until additional material is found.
Although several features of the type suggest emydine characters, a re-examination of the specimen and the reconstructed shell (fig. 25) leads me to consider it to be a valid testudinine species, though actual proof of such a relationship is certainly lacking. Its relationships might eventually be found to be close to Stylemys.
Geochelone incisa (Hay) of the southern United States is certainly closely related to G. osborniana (Hay) and G. orthopygia (Cope) and is here considered to be part of the subgenus Hesperotestudo. The subgenus is best characterized by the presence of a caudal buckler. This structure makes its first appearance in G. amphithorax (Cope) of the Lower Oligocene of Colorado (fig. 26). Two phyletic lines can be recognized in the subgenus.
The Osborniana line: G. brontops (Marsh) of the South Dakota Chadronian seems closely related to G. amphithorax and is probably a member of the subgenus Hesperotestudo.
Figure 26. Geochelone amphithorax (after Hay, 1908).

In these earliest known members of the Osborniana line the shell is relatively thick. The xiphiplastral notch and epiplastral beak are not well developed, being similar to these structures in the subgenus Caudochelys. The supracaudal buckler is small and is composed of relatively few, simple ossicles.
Figure 27. Type of G. osborniana (AMNH 5868).
Known Miocene members of the Osborniana line are G. osborniana (fig. 27) and G. impensa (Hay). In both the shell is larger and proportionally thinner than in the Oligocene species. The supracaudal buckler, and the xiphi- and epi-plastra are similar to those of related
Figure 28. Young specimen of G. osborniana (AMNH 5871).

Figure 29. Adult specimen of G. orthopygia (after Hay, 1908).
forms from the preceding epoch. G. angusticeps (Matthew) is a smaller but more rugose Miocene species that appears to be close to the Pliocene G. orthopygia.
The last member of the Osborniana line is G. orthopygia (figs. 28, 29) of the Lower Pliocene. This is a large, thin-shelled species with a large but simple supracaudal buckler. It represents the culmination of a phyletic line, which, since the Oligocene, tended toward moderate gigantism, a smooth thin shell, and a simple supracaudal buckler.
The Turgida line: Geochelone turgida (Cope, 1892) of the Pliocene of the Great Plains represents the extreme specialization of this phyletic line. It is small and highly rugose. The supracaudal buckler is proportionally large, but not so rugose as that in some Pleistocene forms. G. kalganensis (Gilmore, 1931) from the Kalgan area, northern China, is obviously closely related to G. turgida on the basis of shell rugosity, size, and sulcal patterns. It represents the only record of the subgenus Hesperotestudo outside continental North America. Unfortunately no stratigraphic data are available with the specimen; it is assumed to be Tertiary in age. Testudo insolitus Matthew and Granger (1923) of the Lower Oligocene of Mongolia also may belong to this line, though the evidence is inconclusive.
In the southeastern United States the only known Pleistocene member of this phyletic line is G. incisa. The line is represented in the Pleistocene of the Great Plains area, but complete shells are not available. Although definite allocations are impossible at present, G. equi-

comes (Hay, 1917) probably belongs to this line and, in all likelihood is the western Pleistocene representative of the subgroup. The type (USNM 10944) is an epiplastron. The general shape of the element, particularly its thickness, is almost identical to that of G. incisa, from which I infer a close relationship. G. wilsoni (Milstead, 1956) of the Pleistocene of Texas (fig. 30) is close to G. incisa. G. johnstoni (Auffenberg, 1962), from City Canyon, near Canyon, Texas, is similar to G. incisa, and closely related to it. It is intermediate between G. turgida and G. incisa, both structurally and-temporally. G. riggsi (Hib-bard) is also intermediate (fig. 31), but much closer to G. turgida.
Figure 31. Paratype of G. riggsi, a species seemingly intermediate between G. turgida and G. incisa.
The TurgidaRiggsiJohnstoniIncisa sequence is one in which several trends seem fairly clear (fig. 32). These trends in structural changes with time from Pliocene to Pleistocene include: 1) increase in size, 2) decrease in proportional shell thickness, 3) decrease in rugosity of shell, and 4) additional complexity of the supracaudal buckler through fusion of a larger number of ossicles and a greater degree of ornamentation on the surface of each fused member.
Possibly this line was present in the eastern United States during the Pliocene, though it has not yet been found there. If it was present,

a second species of Pliocene Geochelone from Florida is masquerading as G. hayi.
The subgenus Caudochelys probably begins somewhere near G. uintensis (Gilmore) of the Upper Eocene of Utah. G. ligonia (Cope, 1873) of the Lower Oligocene of Colorado is probably a member of this subgenus. G. arenwaga (Hay, 1907) of the Nebraska Miocene Harrison formation may belong here as a Plains representative, and G. tedwhitei (Williams, 1953), Lower Miocene of Florida, is probably its eastern counterpart. G. hayi (Sellards, 1916), Pliocene of Florida, and G. rexroadensis (Oelrich, 1953), Pliocene of Kansas, may be east-west representatives, though the former is probably more distantly removed from the main line. The latter is certainly closely related to several Pleistocene forms such as G. crassiscutata of Florida. Western Pleistocene counterparts may include G. laticaudata (Cope, 1889).
Relationships within the subgenus Caudochelys seem more obscure than those in Hesperotestudo, perhaps because of the basic similarity of its several known forms. No major trends other than increase in size are recognized.
Figure 33. Type of G. primaeva (UMMP 25758).

In addition to the species discussed above, several other tortoises are known only from remains so fragmentary, or of a type so inadequate for comparison that their relationships remain obscure. Such little-known Pleistocene species are: G. annae (Hay, 1923), and G. francisi (Hay, 1923). Testudo munda Hay (1920) is based on portions of the carapace and plastron of a small turtle (USNM 8944) collected by Ira Sayles in 1885 in a Pleistocene deposit near Whitesburg, Hamblen County, Tennessee. The type is not a testudinine at all, but an emydine. The specimen is, in fact, part of an individual of Terrapene Carolina, and the name T. munda should be added to the synonymy of that species. G. gilberti (Hay, 1902) is an obscure Pliocene form. The little-known Miocene forms are: G. farri (Hay, 1908), G. ducatelli (Collins and Lynn, 1936), G. undata (Cope, 1884), and G. klettiana (Cope, 1874). Oligocene forms are: G. quadrata (Cope, 1884), G. peragrans (Hay, 1906), G. cultrata (Cope, 1873), G. exornata (Lambe, 1913) and G. thompsoni (Hay, 1908). A testudinine turtle described as Testudo primaeva Oelrich (1950), from the Madison River area, Montana, is apparently a primitive form (fig. 33) probably distantly related to Caudochelys and Hesperotestudo. The shell is similar to that of Gopherus emilae (Hay, 1907) from the Rosebud of South Dakota, including the presence of a complete set of hexagonal neurals. Gopherus copei (Koerner, 1940) from the Deep River Miocene of Montana is also similar, though it differs in several minor respects.
Geochelone hayi (Sellards, 1916) may belong to the genus Gopherus, as Williams (1952) suggests. Better preserved specimens should settle this point easily.
Several problems are suggested by the separation of Hesperotestudo of Williams into two subgenera. Both lines apparently arose from Hadrianus, or a Hadrianus-like ancestor. Thus, further study may show the subgenus Hadrianus to be composed of several subgroups. One of these may be ancestral to Geochelone. Perhaps lines ancestral to Hesperotestudo and to Caudochelys may be recognized. Another group may be ancestral to Gopherus, still another to Stylemys.
The last genus presents particularly interesting problems, e.g. 1) the apparent relationship between this genus and Gopherus, and 2) a comparison of the Old and New World forms. Unfortunately, no nonshell elements are known in Floridemys nanus. The proportions of the shell are markedly similar to those of Stylemys, particularly S. conspecta Hay (1908) from the John Day beds of Oregon. In Floridemys the third neural is quadrangular, as it is in many specimens of Stylemys, most Geochelone, and no known Gopherus. Unfortunately

the single specimen of Floridemys is fractured, and it is impossible to ascertain the shape of the second neural. In Stylemys conspecta it is octagonal. The only major difference between them is the peculiar transverse position of the gulohumeral sulcus in Floridemys. S. conspecta is the most Testudo~\ike in the arrangement and proportions of the bony elements of the shell. The most recent, and thus most interesting species of Stylemys in this regard is S. calaverensis Sinclair from the Miocene of California. Unfortunately the fragmentary type is the only specimen. It differs from other species in possessing a first neural that is as broad as the anterior edge of the second, a character also found in Floridemys nanus.
Supracaudal dermal armor, to my knowledge, has not been found with any North American Eocene testudinine. Thus, little can be said regarding the ancestry of the remarkable caudal buckler in Hesperotestudo. Of possible significance is the genus Cymotholcus. Two North American species are recognized: Cymotholcus longus Clark (1932) and C. schucherti (Hay, 1902). The latter is not well known, but the former is represented by a shell and parts of the skeleton. Its relationship to other Eocene testudinines remains obscure, though Clark suggests a possible relationship to Stylemys. Of particular interest in a consideration of the ancestry of Hesperotestudo is the tail of C. longus. According to Clark,
.. Maximum downward projection of the tail would not suffice to allow the tip to touch the plastron. Moreover, because the anal portion of the posterior lobe of the plastron is contained within the floor of the carapace, rather than being ventral to it, the skin of the dorsal region must have stretched horizontally so that the external tail was at most a tiny projection, and very possibly was not present at all.
Though the tail of C. longus is very short, the transverse processes of the vertebrae are not developed to support a buckler. A single dermal ossicle of the caudal area, which was associated with the specimen, could not have been part of such a fused unit. However, in looking for a presumed ancestral form for Hesperotestudo, a short, or even nonexistent external tail is important. Thus, Cymotholcus may be an ancestor to the Hesperotestudo line, although this possibility is suggested now only as an undocumented idea.

Auffenberg, W.
1962. A new species of Geochelone from the Pleistocene of Texas. Copeia, 1962:627-636.
Clarke, J.
1932. A new turtle from the Duchesne Oligocene of the Uinta Basin, northeastern Utah. Ann. Carnegie Mus., vol. 21, no. 3, pp. 131-160.
Collins, R. L., and W. G. Lynn
1936. Fossil turtles from Maryland. Proc. Amer. Philos. Soc, vol. 76, no. 2, pp. 151-173.
Cope, E. D.
1873. Synopsis of new Vertebrata from the Tertiary of Colorado. Govt. Print. Off. Washington, 1873, pp. 1-19.
1874. Report on the vertebrate paleontology of Colorado. 7th Ann. Rept., U.S. Geol. Surv. Terr., pp. 73-86.
1884. The Vertebrata of the Tertiary formations of the West. U.S. Geol.
Surv. Terr., vol. 3, pp. 1-1436. 1889. The vertebrate fauna of the Equus beds. Amer. Nat., vol. 23, pp.
1892. The vertebrate fauna of the Blanco beds. 4th Ann. Rept., U.S. Geol. Surv. Terr., pp. 47-74.
Gilmore, C. W.
1931. Fossil turtles of Mongolia. Bull. Amer. Mus. Nat. Hist., vol. 59, pp. 213-257.
Hay, O. P.
1902. Descriptions of two new species of tortoises from the Tertiary of the United States. Proc. U.S. Natl. Mus., vol. 22, no. 1181, pp. 21-24.
1906. Descriptions of two new genera and two new species of fossil turtles. Bull. Amer. Mus. Nat. Hist., vol. 32, pp. 11-35.
1907. Descriptions of seven new species of turtles from the Tertiary of the United States. Ibid., vol. 32, pp. 847-863.
1908. Fossil turtles of North America. Publ. Carnegie Inst. Washington, no. 75, pp. 1-568.
1916. Descriptions of some Floridian fossil vertebrates belonging mostly to the Pleistocene. Rept. Florida Geol. Surv., vol. 8, pp. 39-76.
1917. On a collection of fossil vertebrates made by Dr. F. W. Cragin in the Equus beds of Kansas. Kansas Univ. Sci. Bull., vol. 10, pp. 39-51.
1920. Descriptions of some Pleistocene vertebrates found in the United
States. Proc. U. S. Natl. Mus., vol. 58, pp. 83-146. 1923. Characteristics of sundry fossil vertebrates. Pan. Amer. Geol., vol. 39,
pp. 101-120.
Holman, J. A.
1959. Amphibians and reptiles from the Pleistocene (Illinoian) of Williston, Florida. Copeia, 1959, pp. 96-102.
Koerner, H. E.
1940. The geology and paleontology of the Fort Logan and Deep River formations of Montana. Amer. lour. Sci., vol. 238, pp. 837-862,

Lambe, L. M.
1913. Description of a new species of Testudo and of a remarkable specimen of Stylemys nebrascensis from the Oligocene of Wyoming. Ottawa Nat., vol. 27, pp. 57-63.
Leidy, J.
1877. Description of vertebrate remains, chiefly from the phosphate beds of South Carolina. Jour. Acad. Nat. Sci. Philadelphia, ser. 2, vol. 8, no. 3, pp. 209-261.
1889. Description of vertebrate remains from Peace Creek, Florida. Trans. Wagner Free Inst. Sci., vol. 2, pp. 13-17. Loomis, F. B.
1927. A giant tortoise from Florida. Amer. Jour. Sci., ser, 4, vol. 28, pp. 17-26.
Loveridge, A., and E. E. Williams
1957. Revision of the African tortoises and turtles of the suborder Cryptodira. Bull. Mus Comp. Zool., vol. 115, no. 6, pp. 163-557.
Matthew, W. D., and W. Granger
1923. The fauna of the Ardyn Obo formation. Amer. Mus. Novitates, no. 98, pp. 1-5. Milstead, W. W.
1956. Fossil turtles of Friesenhahn Cave, Texas, with the description of a new species of Testudo. Copeia, 1956, pp. 161-171.
Oelrich, T. M.
1950. A new Testudo from Madison County, Montana. Contrib. Mus. Paleon-
tol., Univ. Michigan, vol. 8, no. 4, pp. 43-58. 1953. A new Testudo from the Upper Pliocene of Kansas, with additional
notes on associated Rexroad mammals. Trans. Kansas Acad. Sci., vol.
55, no. 3, pp. 300-311.
1957. The status of Upper Pliocene turtle, Testudo turgida Cope. Jour. Pale-ontol, vol. 31, no. 1, pp. 228-241.
Ray, C. E.
1957. A list, bibliography, and index of the fossil vertebrates of Florida. Florida Geol. Surv., Spec. Publ., no. 3, pp. 1-175.
Sellards, E. H.
1916. A new tortoise and a supplementary note on the gavial, Tomistoma americana. Amer. Jour. Sci., ser. 4, vol. 42, pp. 235-240.
Wark, A. F.
1929. A new giant tortoise from the Pliocene of Florida. Amer. Jour. Sci., ser. 5, vol. 17, pp. 400-402. Williams, E. E.
1950. Testudo cuhensis and the evolution of Western Hemisphere tortoises. Bull. Amer. Mus. Nat. Hist., vol. 91, no. 1, pp. 1-36.
1952. A new fossil tortoise from Mona Island, West Indies, and a tentative arrangement of the tortoises of the World. Ibid., vol. 99, no. 9, pp. 541-560.
1953. A new fossil tortoise from the Thomas Farm Miocene of Florida. Bull. Mus. Comp. Zool., vol. 107, no. 11, pp. 537-554.

Contributions to the bulletin of the Florida state museum may be in any field of biology. Manuscripts dealing with natural history or systematic problems involving the southeastern United States or the Caribbean area are solicited especially.
Manuscripts should be of medium length50 to 200 pages. Examination for suitability is made by an Editorial Board.
The bulletin is distributed worldwide through institutional subscriptions and
exchanges only. It is considered the responsibility of the author to distribute his
paper to all interested individuals. To aid in this, fifty copies are furnished the
author without cost, t
Highly recommended as a guide is the volume:
Conference of Biological Editors, Committee on Form and Style. 1960. Style manual for biological journals.
Amer. Inst. Biol. Sci., Washington. 92 p.
Manuscripts shoidd be typewritten with double spacing throughout, with ample margins, and on only one side of the paper. The author should keep a copy; the original and a carbon must be submitted. Tables and legends of figures should be typed on sheets separate from the text. Several legends or tables may be placed on a single sheet.
Illustrations, including maps and photographs, should be referred to as "figures." All illustrations are reduced to a maximum of 4-1/4 by 7-1/8 inches. Size scales, wherever they are necessary, should be incorporated into the figure.
References to literature should conform with the bibliographic style used in recent numbers of the bulletin. Spell out in full the titles of non-English serials and places of publication.
Footnote material should be kept to a minimum. However, provide copy for a footnote detailing the title, affiliations, and address of the author (see recent numbers of the bulletin).
Manuscripts must be accompanied by a synopsisa brief and factual summary (not a mere description) of the contents and conclusions, which points out the presence of any new information and indicates its relevance. In it list all new organisms described and give their ranges; indicate all taxonomic changes proposed. The synopsis, written in full sentences, should be concise, but completely intelligible in itself without references to the paper, thereby enabling the busy reader to decide more surely than he can from the title alone whether the paper merits his reading. The synopsis will be published with the paper. It does not replace the usual conclusions or summary sections. It may also serve as copy for the abstracting services.
Manuscripts and all editorial matters should be addressed to:
Managing Editor of the bulletin Florida State Museum Seagle Building Gainesville, Florida


University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs