550 5 GEOLOGY LIBRARY
r iELDIANA
Geology
NEW SERIES, NO. 45
The Cranial Anatomy of Placochelys placodonta Jaekel, 1902, and a Review of the Cyamodontoidea (Reptilia, Placodonta)
Olivier Rieppel
£J October 31, 2001
{== Publication 1514
CO
Z PUBLISHED BY FIELD MUSEUM OF NATURAL HISTORY
«EWLUB
at iirbANA-CHAMPAIGN
HI un QpQLOQYlnfermation for Contributors to Fieldiana
Generall|Jjk/zvL is primarily a journal for Field Museum staff members and research associates, although manuscripts fromnonaffiliated authors may be considered as space permits.
The Journal carries a page charge of $65.00 per printed page or fraction thereof. Payment of at least 50 /o of page charges qualifies a paper for expedited processing, which reduces the publication time. Contributions from staff, research associates, and invited authors will be considered for publication regardless of ability to pay page charges however the full charge is mandatory for nonaffiliated authors of unsolicited manuscripts. Three complete copies oi the text '(including title page and abstract) and of the illustrations should be submitted (one original copy plus two review copies which may be machine, copies). No manuscripts will be considered for publication or submitted to reviewers before all materials are complete and in the hands of the Scientific Editor.
Manuscripts should be submitted to Scientific Editor, Fieldiana, Field Museum of Natural History, Chicago, Illinois 60605-2496, U.S.A.
Text- Manuscripts must be typewritten double-spaced on standard-weight, 8%- by 11 -inch paper with wide margins on all four sides. If typed on an IBM-compatible computer using MS-DOS, also submit text .on 5%-inch diskette (WordPerfect 4.1, 4.2, or 5.0, MultiMate, Displaywrite 2, 3 & 4, Wang PC, Samna, Microsoft Word, Volks- writer, or WordStar programs or ASCII). n
For papers over 100 manuscript pages, authors are requested to submit a "Table of Contents, a List ot Illustrations," and a "List of Tables" immediately following title page. In most cases, the text should be preceded by an "Abstract" and should conclude with "Acknowledgments" (if any) and "Literature Cited."
All measurements should be in the metric system (periods are not used after abbreviated measurements). The format and style of headings should follow that of recent issues of Fieldiana.
For more detailed style information, see The Chicago Manual of Style (13th ed.), published by The University of Chicago Press, and also recent issues of Fieldiana.
References- In "Literature Cited," book and journal titles should be given in full. Where abbreviations are desirable (e g in citation of synonymies), authors consistently should follow Botanico-Periodicum-Huntianum and TL-2 Taxonomic Literature by F. A. Stafleu & R. S. Cowan (1976 et seq.) (botanical papers) or Serial Sources for the Biosis Data Base (1983) published by the BioSciences Information Service. Names of botanical authors should follow the "Draft Index of Author Abbreviations, Royal Botanic Gardens, Kew," 1984 edition, or TL-2. References should be typed in the following form:
Croat T B 1978 Flora of Barro Colorado Island. Stanford University Press, Stanford, Calif, 943 pp. Grubb P. J J R Lloyd, and T D. Pennington. 1963. A comparison of montane and lowland rain forest in
' Ecuador I. The forest structure, physiognomy, and floristics. Journal of Ecology, 51: 567-601. Langdon E J. M. 1979. Yage among the Siona: Cultural patterns in visions, pp. 63-80. In Browman, D. L.,
and R A Schwarz, eds., Spirits, Shamans, and Stars. Mouton Publishers, The Hague, Netherlands Murra J 1946.' The historic tribes of Ecuador, pp. 785-821. In Steward, J. H., ed Handbook of South American Indians. Vol. 2, The Andean Civilizations. Bulletin 143, Bureau of American Ethnology, Smithsonian Institution, Washington, D.C. Stolze, R. G. 1981. Ferns and fern allies of Guatemala. Part II. Polypodiaceae. Fieldiana: Botany, n.s., 6: 1-
522. Illustrations: Illustrations are referred to as "figures" in the text (not as "plates"). Figures must be accompanied by some indication of scale, normally a reference bar. Statements in figure captions alone, such as "X0.8 are not acceptable. Captions should be typed double-spaced and consecutively. See recent issues of Fieldiana for details of
All illustrations should be marked on the reverse with author's name, figure number(s), and "top."
Figures as submitted should, whenever practicable, be 8V2 by 1 1 inches (22 X 28 cm) and may not exceed 1 IV* by I6V2 inches (30 X 42 cm). Illustrations should be mounted on boards in the arrangement to be obtained in the printed work This original set should be suitable for transmission to the printer as follows: Pen and ink drawings may be originals (preferred) or photostats; shaded drawings must be originals, but within the size limitation; and photostats must be high-quality, glossy, black and white prints. Original illustrations will be returned to the corre- sponding author upon publication unless otherwise specified.
Authors who wish to publish figures that require costly special paper or color reproduction must make prior arrangements with the Scientific Editor.
Page Proofs: Fieldiana employs a two-step correction system. The corresponding author will normally receive a copy of the edited manuscript on which deletions, additions, and changes can be made and queries answered. Only one set of page proofs will be sent. All desired corrections of type must be made on the single set of page proofs. Changes in page proofs (as opposed to corrections) are very expensive. Author-generated changes in page proofs can only be made if the author agrees in advance to pay for them.
© This paper meets the requirements ot ANSI/NISO Z39.48-1992 (Permanence ot Paper).
FIELDIANA
Geology
NEW SERIES, NO. 45
The Cranial Anatomy of Placochelys placodonta Jaekel, 1902, and a Review of the Cyamodontoidea < Kept i I in, Placodonta)
Olivier Rieppel
Department of Geology
Field Museum of Natural History
1400 South Lake Shore Drive
Chicago, Illinois 60605-2496
U.S.A.
Accepted March 7, 2000 Published October 31, 2001 Publication 1514
PUBLISHED BY FIELD MUSEUM OF NATURAL HISTORY
© 2001 Field Museum of Natural History
ISSN 0096-265 1
PRINTED IN THE UNITED STATES OF AMERICA
no
Table of Contents
«-VU/UTUfcfft*ft,
u-r
Abstract 1
Introduction 1
Material Included in This Study 2
List of Abbreviations Used in the Figures
3
The Cranial Anatomy of Placochelys pla-
codonta 3
Measurements of the Holotype 5
Morphological Description of the Skull 6
Morphological Description of the Lower
Jaw 21
Dermal Ornamentation of the Skull 23
Comparison of the Cranial Anatomy of Placochelys placodonta with That of
Other Cyamodontoid Placodonts 23
The Cranial Anatomy of Cyamodus ros-
tratus (Minister, 1839) 24
The Cranial Anatomy of Cyamodus
muensteri (Agassiz, 1839) 33
The Cranial Anatomy of Cyamodus "/a-
ticeps" (Owen, 1858) 35
The Cranial Anatomy of Cyamodus kuhnschnyderi Nosotti and Pinna,
1993a 37
The Cranial Anatomy of Cyamodus hilde-
gardis Peyer, 1931a 39
The Cranial Anatomy of Henodus chel-
yops v. Huene, 1936 39
The Cranial Anatomy of Macroplacus
raeticus Schubert-Klempnauer, 1975 .. 49 The Cranial Anatomy of Protenodonto-
saurus italicus Pinna, 1990b 56
The Cranial Anatomy of Psephoderma
alpinum H. v. Meyer, 1858 59
Autapomorphies in the Skull of the Cy-
amodontoidea 63
Evolution of the Rostrum and of the Den- tition Within the Cyamodontoidea 65
Cladistic Analysis of Cyamodontoid In- terrelationships 67
Systematic Paleontology 71
Cyamodontoidea Nopcsa, 1923 71
Cyamodontida, new taxon 72
Henodus Huene, 1936 72
Henodus chelyops Huene, 1936 72
Cyamodontidae Nopcsa, 1923 73
Cyamodus Meyer, 1863 73
Cyamodus hildegardis Peyer, 1931a 73
Cyamodus kuhnschnyderi Nosotti and
Pinna, 1993 74
Cyamodus muensteri (Agassiz, 1839) 75
Cyamodus rostratus (Munster, 1839) 76
Placochelyida, new taxon 78
Macroplacus Schubert-Klempnauer, 1975 .. 78 Macroplacus raeticus Schubert-Klemp- nauer, 1975 78
Unnamed Taxon 78
Protenodontosaurus Pinna, 1990b 78
Protenodontosaurus italicus Pinna, 1990b .. 79
Placochelyidae Romer, 1956 79
Placochelys Jaekel, 1902 79
Placochelys placodonta Jaekel, 1 902 79
Psephoderma Meyer, 1858 85
Psephoderma alpinum Meyer, 1858 86
Placochelys and Potential Turtle Rela- tionships of the Cyamodontoidea 87
Paleobiogeography and Paleoecology of
Cyamodontoid Placodonts 92
Acknowledgments 95
Literature Cited 95
Appendix I 101
List of Illustrations
1 . Skull of Placochelys placodonta Jaekel (holotype) 4
2. Skull of Placochelys placodonta Jaekel (paratype) 5
3. Skull of Placochelys placodonta Jaekel (holotype) 8
4. Skull of Placochelys placodonta Jaekel (holotype) 10
5. Skull of Placochelys placodonta Jaekel (paratype) 12
6. Braincase of Placochelys placodonta Jaekel (holotype) 14
7. Occiput of Placochelys placodonta Jae- kel (paratype) 18
8. Occiput of Placochelys placodonta Jae- kel (holotype) 19
9. Lower jaw of Placochelys placodonta Jaekel (holotype) 22
10. Lower jaw of Cyamodus hildegardis
Peyer 23
1 1 . Skull of Cyamodus rostratus Munster (holotype) 24
12. Skull of Cyamodus rostratus Munster (holotype) 26
13. Skull of Cyamodus rostratus Munster (holotype and referred specimen) 28
14. Pterygoid of Cyamodus rostratus Mun- ster (referred specimen) 31
in
15. Lower jaw of Cyamodus rostratus Miinster (referred specimen) 32
16. Skull of Cyamodus muensteri (holo-
type) 33
17. Skull of Cyamodus muensteri (holotype
of C. "laticeps" Owen) 34
18. Skull of Henodus chelyops v. Huene (syntype) 40
19. Skull of Henodus chelyops v. Huene (syntype and referred specimen) 41
20. Skull of Henodus chelyops v. Huene (syntypes) 42
2 1 . Skull of Henodus chelyops v. Huene (syntype and referred specimen) 44
22. Premaxillary denticles in Henodus chelyops v. Huene 44
23. Right side of dermal palate in Henodus chelyops v. Huene 46
24. Suspension of left quadrate in Henodus chelyops v. Huene 47
25. Lower jaw of Henodus chelyops v.
Huene 48
26. Skull of Macroplacus raeticus Schu- bert-Klempnauer (holotype) 50
27. Skull of Protenodontosaurus italicus
Pinna (holotype) 54
28. Skull of Psephoderma alpinum H. v. Meyer 59
29. Skull of Psephoderma alpinum H. v. Meyer 60
30. Left lateral braincase wall of Placodus gigas Agassiz 64
31. Most parsimonious unrooted network
for Placodontoidea 67
32. Most parsimonious reconstruction of placodont interrelationships 68
33. Most parsimonious reconstruction of placodont interrelationships, with Cy- amodus hildegardis included 69
34. Strict consensus tree for placodont in- terrelationships, with the Negev speci- men included 71
35. Skull of Placochelys alpis sordidae
Broili (holotype) 81
36. Skull of Placochelys alpis sordidae
Broili (holotype) 81
37. Holotype of Placochelys stoppanii Oss- wald 83
38. Dermal ossifications referred to IPla- cochelys 84
39. Area cladogram for the cyamodontoid placodonts 94
List of Tables
1 . Measurements of maxillary and pala- tine tooth plates of Placochelys placo- donta (holotype) 5
2. Measurements of the dentary tooth plates of Placochelys placodonta (holo- type) 21
3. Measurements of the palatine tooth plates of Cyamodus rostratus (holo- type) 29
4. Measurements of the dentary tooth plates of Cyamodus rostratus (referred specimen) 32
5. Dentitional characters of species of the genus Cyamodus 33
6. Measurements of the tooth plates of Cyamodus muensteri 37
7. Proportions of the posterior palatine tooth plate throughout the Cyamodon- toidea 39
8. Skull proportions of cyamodontoid pla- codonts 51
9. Measurements of the maxillary and palatine tooth plates of Macroplacus raeticus (holotype) 53
10. Measurements of the maxillary and palatine tooth plates of Protenodonto- saurus italicus (holotype) 58
1 1 . Data matrix for the analysis of placo- dont relationships 66
12. Measurements of the maxillary and palatine tooth plates of Placochelys mal- anchinii Boni, 1947 84
13. Measurements of the maxillary and palatine tooth plates of Psephoderma alpinum 85
14. Data matrix for the implementation of the Brooks parsimony analysis in the reconstruction of cyamodontoid histori- cal biogeography 85
IV
The Cranial Anatomy of Placochelys placodonta Jaekel, 1902, and a Review of the Cyamodontoidea (Reptilia, Placodonta)
Olivier Rieppel
Abstract
The skull of Placochelys placodonta Jaekel is described in detail and compared with all other cyamodontoid skulls kept in public repositories. Cladistic analysis based on a character set derived from cyamodontoid skull anatomy results in a reconstruction of placodont interre- lationships as follows: (Paraplacodus (Placodus ((Henodus, Cyamodus) (Macroplacus (Pro- tenodontosaurus (Placochelys, Psephoderma)))))). The monophyly of the Cyamodontoidea is very robust, supported in particular by a suite of derived braincase characters. On the basis of present evidence, Henodus is the sister taxon of Cyamodus, and the monophyletic genus Cy- amodus includes C. hildegardis. The monophyly of placochelyids, including Placochelys and Psephoderma, is strongly supported also.
A detailed comparison of skull anatomy provides no basis for a hypothesis of close phylo- genetic relationships of turtles and cyamodontoid placodonts. Any similarities between the two clades, particularly with respect to the development of extensive dermal armor, must be con- vergent.
The historical paleobiogeography of cyamodontoid placodonts can largely be understood as a sequence of vicariance events that involved an early bifurcation establishing separate clades in the Germanic Basin and on the Eurasian carbonate platform. Subsequent vicariance estab- lished separate clades in the northern Alpine Triassic and in the southern Alps on the Hungarian platform, with further subdivision of the clades within the latter.
Introduction
Cyamodontoid placodonts are a clade of marine reptiles that occurs in shallow epicontinental and nearshore deposits of Middle and Upper Triassic age throughout the western Tethyan faunal prov- ince. The earliest cyamodontoid placodont to ap- pear in the fossil record is Cyamodus tarnowitz- ensis Giirich, 1884, a skull (now lost) from the Karchowice Beds of Tarnowskie Gory, Poland (formerly Tarnowitz in Upper Silesia), which be- long to the uppermost lower Muschelkalk (lower Illyrian, lower Anisian). Another early occurrence is a skull fragment (Brotzen, 1957; Rieppel, Ma- zin, & Tchernov, 1999) of a cyamodontoid from the lower Muschelkalk of Makhtesh Ramon, Ne- gev, Israel (Middle Member of the Gevanim For-
mation, upper Bithynian, upper lower Anisian: Druckman, 1974). The latest occurrences of cy- amodontoids are from the Rhaetian of the north- ern (Psephoderma: Meyer, 1858a, b; Broili, 1921; Macroplacus: Schubert-Klempnauer, 1975) and southern (Psephoderma: Osswald, 1930; Boni, 1946 [1947], 1947 [1948]; Pinna, 1975, 1976a, b, 1978, 1979; Pinna & Nosotti, 1989) Alpine Tri- assic and from the Rhaetian of England (Meyer, 1858a, b; Storrs, 1994; see also Pinna, 1990a). Other localities that have yielded cyamodontoid placodonts are in the Anisian of Transylvania (Jurcsak, 1982; Huza et al., 1987); the upper Mu- schelkalk and Keuper of southern Germany (Cy- amodus rostratus, Cyamodus muensteri, and Cy- amodus "laticeps" from the upper Anisian [Ag- assiz, 1833-45; Minister, 1839; Owen, 1858;
FIELDIANA: GEOLOGY, N.S., NO. 45, OCTOBER 31, 2001, PP. 1-104
Meyer, 1863 J; Cyamodus kuhnschnyderi from the lower Ladinian [Nosotti & Pinna, 1993a]; Pse- phosaurus suevicus from the upper Ladinian [Fraas, 1896]; Henodus chelyops from the Carni- an [Huene, 1936]) and of the Lorraine, France (Corroy, 1928; Rieppel & Hagdorn, 1999); the Ladinian of the southern Alps {Cyamodus hilde- gardis, Peyer, 1931a); the middle Carnian of the Tre Venezie area of northeastern Italy (Pinna & Zucchi Stolfa, 1979; Dalla Vecchia, 1993; Proten- odontosaurus italicus, Pinna, 1990b); the Ladini- an of northeastern Spain (Rieppel and Hagdorn, 1998); and Middle Triassic (?Anisian, Ladinian) localities on the northern Gondwanan shelf (Haas, 1959, 1975; Gorce, 1960; Beltan et al., 1979; Vickers-Rich et al., 1999).
Cyamodontoid placodonts were a widespread and taxonomically diverse group characterized by the development of extensive dermal armor, which enhanced their chances of representation in the fossil record. In its most derived condition, this dermal armor consisted of a solid carapace, linked to a ventral armor by a lateral dermal body wall (Haas, 1969). This resulted in a remarkably turtle-like appearance of cyamodontoid placo- donts, so much so that, based on his study of the cranial anatomy and dermal armor of Placochelys placodonta, Jaekel (1902a, b, 1907) proposed a derivation of turtles from cyamodontoids. The hy- pothesis of a relationship of turtles to placodonts was later rejected by Gregory (1946), who noted that convergent evolution is remarkable in these two groups, especially with regard to the dermal armor.
The more recent finding that turtles may be the sister-group of Sauropterygia among crown-group Diapsida (Rieppel & deBraga, 1996; deBraga & Rieppel, 1997; Rieppel & Reisz, 1999) has brought cyamodontoid placodonts back into fo- cus. Although a broad-based analysis of turtle re- lationships over a wide range of taxa confirmed that the similarities shared by turtles and cyamo- dontoids are convergent (Rieppel & Reisz, 1999), a more in-depth analysis of the cranial anatomy of cyamodontoid placodonts and its comparison with that of turtles appears desirable in the at- tempt to discover further similarities or differen- ces between the two groups. Placochelys placo- donta was selected as primary focus for this pro- ject not only because Jaekel (1902a, b, 1907) based his hypothesis of a turtle-placodont rela- tionship on this taxon, but also because it repre- sents one of the best-preserved cyamodontoid skulls. The cranial anatomy of Placochelys pla-
codonta will also be compared in detail with the cranial anatomy of all other cyamodontoids for which skull material is available in an effort to analyze the phylogenetic interrelationships within the Cyamodontoidea. This will provide the nec- essary framework for the identification of the bas- al cyamodontoid skull morphology and the anal- ysis of evolutionary changes of skull morphology within this group of fascinating reptiles.
Material Included in This Study
The following is a list of material included in the present study. Institutional abbreviations are: bmnh: British Museum (Natural History), now The Natural History Museum, London; bsp, Bay- erische Staatssammlung fiir Palaontologie und historische Geologie, Munich; fafi, Magayar Al- lami Foldtani Intezet (Geological Institute of Hun- gary, Budapest); gpit, Geologisch-Palaontolo- gisches Institut, Universitat Tubingen; HUJ-Pal., Paleontological Collections, Department of Evo- lution, Systematics and Ecology, Hebrew Univer- sity, Jerusalem; mb.r., Museum fiir Naturkunde der Humboldt Universitat, Berlin, fossil reptile collection; mbsn, Museo Brembano di Scienze Naturali, San Pellegrino; mfsn, Museo Friulano di Storia Naturale, Udine; msnb, Museo Civico di Scienze Naturali "E. Caffi," Bergamo; msnm, Museo Civico di Storia Naturale di Milano; pimuz, Palaontologisches Institut und Museum der Uni- versitat Zurich; smf, Senckenberg Museum, Frankfurt a.M.; smns, Staatliches Museum fiir Na- turkunde, Stuttgart; st, Museo della Vicaria di S. Lorenzio, Zogno (Bergamo, Italy); umo, Urwelt- Museum Oberfranken, Bayreuth.
Cyamodus hildegardis Peyer, 1931a: pimuz T4763 (holotype), T4768 (original of Peyer, 1935, PI. 46, Figs, la-c, and Pinna, 1992, Fig. 6), T4771 (original of Pinna, 1992, Fig. 7), T2796 (original of Kuhn-Schnyder, 1959, PI. I, and Pinna, 1992, Fig. 8).
Cyamodus "laticeps" (Owen, 1858): bmnh R 1644 (holotype).
Cyamodus kuhnschnyderi Nosotti and Pinna, 1993a: smns 15855 (holotype), smns 16270 (par- atype); mhi 1294 (incomplete skull).
Cyamodus muensteri (Agassiz, 1839): bsp AS VII 1210 (holotype, original of Minister, 1830, skull no. II; Meyer, 1863, PI. 31, Figs. 1-2).
Cyamodus rostratus (Munster, 1839): umo BT
FIELDIANA: GEOLOGY
748 (holotype, original of Drevermann, 1928; Kuhn-Schnyder, 1965a).
Cyamodus cf. rostratus: smns 17403 (incom- plete skull, original of Nosotti and Pinna, 1993b, Fig. 3); umo BT 2172 (isolated lower jaw, original of Drevermann, 1928, PI. 23, Fig. 2); smf R-4040 (isolated lower jaw, original of Drevermann, 1928, PI. 23, Figs. 3a-d, and of Rieppel, 1995a, Fig. 31).
Henodus chelyops v. Huene, 1936: "Specimens I and II," syntypes of Huene (1936); "specimens IV and VI," collected in 1959 (Fischer, 1959). All specimens are kept at the gpit, uncatalogued.
Macroplacus raeticus Schubert- Klempnauer, 1975: bsp 1967 I 324 (holotype).
Placochelyanus malanchinii Boni, 1947 (1998): msnm uncatalogued (cast of holotype).
Placochelyanus stoppanii Osswald, 1930: bsp AS I 1457 (holotype).
Placochelys alpis sordidae Broili, 1921: bsp 1921.1.3 (holotype).
Placochelys placodonta Jaekel, 1902b: fafi Ob/ 2323/Vt.3. (holotype); mb.r. 1765 (paratype).
Placodus gigas Agassiz, 1839: umo Bt 13 (skull, original of Sues, 1987, and Rieppel, 1995a).
Protenodontosaurus italicus Pinna, 1990b: mfsn 1819GP (holotype), mfsn 1923GP (referred second specimen).
Psephoderma alpinum H. v. Meyer, 1858: bsp AS I 8 (holotype, carapace); msnm V471 (skull; referred specimen); msnb 4884a-b (juvenile skull, original of Pinna, 1979).
Psephosaurus mosis Brotzen, 1957: HUJ-Pal. 220 (referred specimen).
|
fcst |
facet (on quadrate) receiving the shaft |
|
of the stapes |
|
|
f-jug |
jugular foramen |
|
f.l |
lacrimal foramen |
|
f.p.dl |
posterior dental lamina foramen |
|
f.pin |
pineal foramen |
|
f.trig |
trigeminal foramen |
|
f.vest |
vestibular (oval) fenestra |
|
in |
internal naris |
|
ju |
jugal |
|
m |
maxilla |
|
nu- |
Meckel's canal |
|
ll |
nasal |
|
op |
opisthotic |
|
P |
parietal |
|
Pi |
palatine |
|
I > in |
premaxilla |
|
po |
postorbital |
|
pof |
postfrontal |
|
pq.rc |
palatoquadrate cartilage recess |
|
pra |
prearticular |
|
prf |
prefrontal |
|
pro |
prootic |
|
pt |
pterygoid |
|
pt.f |
posttemporal fenestra |
|
pto.f |
pteroccipital foramen |
|
q |
quadrate |
|
qj |
quadratojugal |
|
sang |
surangular |
|
so |
supraoccipital |
|
sp |
splenial |
|
sq |
squamosal |
|
sq.bt |
squamosal buttress (receiving the distal |
|
end of the paroccipital process) |
|
|
stc |
sella turcica |
|
V |
vomer |
|
List of Abbreviations Used in the |
|
|
Figures |
|
|
ang |
angular |
|
ar |
articular |
|
bo |
basioccipital |
|
c |
coronoid |
|
cci |
canal for internal carotid |
|
d |
dentary |
|
ds |
dorsum sellae |
|
eo |
exoccipital |
|
ep |
epipterygoid |
|
ep.o |
epiotic ossification |
|
f |
frontal |
|
f.cc |
foramen for cerebral carotid |
|
f.ch.t. |
chorda tympani foramen |
The Cranial Anatomy of Placochelys placodonta
Remains of Placochelys placodonta were first collected in 1899 by Desiderius Laczk6 in the Also Keuper (upper Middle Triassic) of Jeruzsa- lemhegy (Jerusalem mountain) near Veszprem, a small town in west central Hungary, located on the south slopes of the Bakony Mountains over- looking Lake Balaton. Subsequent collecting ef- forts yielded two skulls, several carapace frag- ments, and scattered remains of the postcranial skeleton. The new genus and species was de- scribed by Jaekel in 1902, who subsequently pre- sented the material in a comprehensive mono-
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
Fig. 1. Skull of Placochelys placodonta Jaekel (holotype, fafi Ob/2323/Vt.3): A, dorsal view; B, ventral view; C, left lateral view. Scale bar = 20 mm.
graph (Jaekel, 1907). The holotype (skull, speci- men I of Jaekel, 1907) is three-dimensionally pre- served (Fig. 1), but was incompletely prepared at the time of its original description. A cast of the holotype in its original condition, that is, as de- scribed by Jaekel (1902a, b, 1907), is kept at the Geological Institute of Hungary. The second skull (Fig. 2) is dorsoventrally compressed and was fig- ured in ventral view only by Jaekel (1907, PI. III). The holotype was later sent to Frankfurt a.M. for further preparation by Christian Strunz, be- cause Fritz Drevermann planned to study the specimen in greater detail. Strunz separated the lower jaw from the cranium and fully exposed the
braincase. Drevermann never got around to de- scribing the specimen, but it was briefly dealt with in a publication by Huene (1931). The figures published by Huene (1931) are erroneous in many details, as will be discussed below. In his mono- graph, Jaekel (1907) made only passing reference to the second skull, which later received no fur- ther attention other than a photograph included in Kuhn-Schnyder (1965b, Fig. 8) and Miiller (1968, Fig. 235).
Parts of the postcranial remains of Placochelys placodonta were lost during World War II (West- phal, 1975). The material now missing comprises the following elements illustrated by Jaekel
FIELDIANA: GEOLOGY
Fig. 2. Skull of Placochelys placodonta Jaekel (paratype, mb.r. 1765): A, dorsal view; B, ventral view. Scale bar = 20 mm.
(1907): PI. V, Figs. 2-7; PI. PI. VI, Fig. 4; PI. VII, Figs. 3, 5-6, 8, 10-11; PI. VIII, Figs. 1-18. Casts of the right femur (bmnh R 4070, 4074) and of the left humerus (bmnh R 4069) are kept at The Natural History Museum, London.
Measurements of the Holotype
The anterior tip of the rostrum is broken in the holotype. All measurements are given as pre-
Table 1. Measurements of maxillary and palatine tooth plates of Placochelys placodonta (holotype, mafi Ob/2323/Vt.3). All measurements in mm; approximate values in parentheses.
|
left |
right |
|||
|
longi- tudinal 0 |
trans- verse 0 |
longi- tudinal 0 |
trans- verse 0 |
|
|
anterior maxillary tooth |
8.2 |
- |
- |
6.7 |
|
intermediate maxillary tooth |
8.9 |
7.4 |
8.9 |
7.2 |
|
posterior maxillary tooth |
15.7 |
11.4 |
- |
11.8 |
|
anterior palatine tooth |
13.8 |
11.0 |
14.0 |
11.0 |
|
posterior palatine tooth |
27.2 |
20.8 |
(26) |
21.0 |
served in the fossil. Measurements in parentheses are those of the right side of the skull. Measure- ments for the maxillary and palatine tooth plates are given in Table 1.
Tip of rostrum to occipital condyle: 1 18.3 mm Tip of rostrum to mandibular condyle of quadrate:
119.2 (116.4) mm Tip of rostrum to posterior margin of skull table:
112.8 mm Maximal length of skull: 148 mm Tip of rostrum to anterior margin of external nar-
is: 21.2 (20.5) mm Tip of rostrum to anterior margin of internal naris:
32.0 (32.0) mm Tip of rostrum to anterior margin of orbit: 39.1
(38.5) mm Tip of rostrum to anterior margin of upper tem- poral fossa: 83.5 (82.1) mm Longitudinal diameter of external naris: 12.5
(13.0) mm Transverse diameter of external naris: 9.8 (10.0)
mm Longitudinal diameter of internal naris: 7.1 (7.0)
mm
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
Transverse diameter of internal naris: 4.3 (4.1)
mm Longitudinal diameter of orbit: 26.1 (27.4) mm Transverse diameter of orbit: 27.4 (24.8) mm Longitudinal diameter of upper temporal fossa:
52.1 (51.5) mm Transverse diameter of upper temporal fossa: 32.9
(32.4) mm Distance from external naris to orbit: 7.0 (7.0)
mm Distance from orbit to upper temporal fossa: 18.5
mm Bony bridge between external nares: 6.3 mm Bony bridge between internal nares: 4.0 mm Bony bridge between orbits: 10.6 mm Bony bridge between upper temporal fossae: 8.2
Morphological Description of the Skull
All cyamodontoid placodonts with the excep- tion of Henodus (Huene, 1936) have a skull with a broadly expanding temporal region and a nar- row, tapering rostrum. Proportions vary from a rather short and blunt rostrum in Cyamodus (No- sotti & Pinna, 1996) to a very elongate, narrow rostrum in Psephoderma (Pinna & Nosotti, 1989; see also further discussion below). Placochelys is intermediate between these two extremes, with an elongate and narrow rostrum formed by the pre- maxillae (Fig. 3A). The tip of the rostrum is bro- ken in the holotype, and the rostrum is altogether missing in mb.r. 1765. The rostrum is sufficiently well preserved in the holotype, however, to allow the conclusion that the premaxilla of Placochelys was edentulous. In ventral view, distinct longitu- dinal grooves can be seen running from the an- terior margins of the external nares anteriorly up to the broken tip of the rostrum. These grooves are delineated laterally by the maxilla (in their posterior part) and by the premaxilla (in their an- terior part), and medially by a raised ventral crest running along the medial edge of each premaxilla.
The premaxilla broadly enters the anterior and lateral margin of the external naris. It forms a dis- tinct (autapomorphic) posterior process that ex- tends backward from the posteroventral corner of the external naris to a level shortly behind the anterior margin of the orbit, embraced both dor- sally and ventrally by the maxilla (Fig. 4A). A comparable process is not observed in other cy- amodontoid taxa. Dorsomedially, the premaxilla
meets the nasal between the two external nares in a posterolaterally trending suture.
In ventral view (Fig. 3B), the contact of the premaxilla with the maxilla posterolaterally and the vomer posteriorly is obscured by paint cov- ering the bone surface. In dorsal view, however, the anterior end of the maxilla can be seen to ex- pand medially to form most of the dermal floor of the external naris. The ventral outline of the anterior end of the maxilla has been reconstructed accordingly (broken lines in Fig. 3B), which in- dicates that the maxilla met the vomer along the anterior margin of the external naris, excluding the premaxilla from the latter.
The maxilla forms a slender anterior process that runs along the ventrolateral margin of the ros- trum to a level well in front of the anterior margin of the external naris (Fig. 4A). It remains sepa- rated from the lateral margin of the external naris by the posterior process of the premaxilla. The maxilla expands medially below and deep to this posterior process of the premaxilla, as it forms most of the floor of the external naris. Between the external naris and the orbit, the maxilla forms a distinct ascending process with a pointed dorsal tip wedged in between the nasal anteriorly and the prefrontal posteriorly. The maxilla narrowly en- ters the anteroventral margin of the orbit, but fur- ther posteriorly it is excluded from the ventral margin of the orbit by the anterior process of the jugal. The posterior end of the maxilla forms an essentially vertical and deeply interdigitating su- ture with the jugal at a level somewhat in front of the posterior margin of the orbit but behind the level of the midpoint of the longitudinal diameter of the orbit. The lateral surface of the maxilla shows five to seven superior labial foramina (Fig. 4A).
In ventral view (Fig. 3B), the maxilla is seen to enter the anterolateral margin of the internal naris. It contacts the vomer anteromedially in front of the external naris (the suture between the two bones is distinct at the anterior margin of the internal naris), and the palatine lateral to the ex- ternal naris. The maxilla remains excluded from the anterior margin of the subtemporal fossa by a lateral process of the palatine, which contacts the jugal. Each maxilla carries three tooth plates, of which the posteriormost one is distinctly larger than the two anterior ones (Table 1). In contrast to the holotype, mb.r. 1765 shows a distinct dental lamina foramen located on the palatine-maxillary suture posteromedial to the posterior maxillary tooth plate.
FIELDIANA: GEOLOGY
The nasals are paired, triangular elements that define the posteromedial margin of the external nares. They meet each other along the dorsal mid- line of the skull, separating the premaxilla from the frontal (Fig. 3A). The anterior tips of the na- sals lie at a level behind the anterior margin of the external nares. In the holotype, a narrow but deep cleft separates the nasals from one another posteriorly, exposing the underlying frontal. Giv- en the overall solid ossification of the skull and the tendency for the sutures to fuse in the dermal skull roof, it seems unlikely that the narrow cleft between the nasals reflects incomplete ossification of these latter elements. In no other cyamodontoid skull are the posterior parts of the nasals separated by a deep cleft exposing the underlying frontal. The posterolateral margin of the nasal runs from the posterior margin of the external naris in a pos- teromedial direction, contacting the ascending process of the maxilla, the anterior margin of the prefrontal, and the short anterolateral process of the frontal. The anterolateral process of the frontal therefore remains separated from the ascending process of the maxilla by the nasal and prefrontal (Fig. 3A).
The prefrontal is a rather small element located at the anterodorsal margin of the orbit. A medial ventral process forms the anteromedial margin of the orbit. The location of the lacrimal foramen cannot be identified unequivocally in Placochelys. The anteroventral corners of the orbits are not pre- served in mb.r. 1765. The anteroventral margin of the right orbit is subject to breakage in the holo- type. In the left orbit of the holotype, the prefron- tal is seen to extend further down than in Cyamo- dus kuhnschnyderi and Protenodontosaurus, where the prefrontal remains excluded from the lacrimal foramen (Nosotti & Pinna, 1996, 1998; see further discussion of the latter two taxa be- low). In Placochelys the medial ventral process of the prefrontal reaches the maxilla and closely ap- proaches the anterior process of the jugal without quite reaching it (Fig. 3A). The position of the lacrimal foramen is again obscured by breakage and compression. However, this break might pass through an area of weakness indicating the posi- tion of the lacrimal foramen, in which case the prefrontal might have entered its dorsal margin. Breakage likewise obscures the location of the fo- ramen for the passage of the infraorbital division of the maxillary branch of the trigeminal nerve (infraorbital foramen sensu Oelrich, 1956) in the anteroventral corner of the orbit. However, Pla- cochelys does not show a distinct groove running
along the anteroventral margin of the orbit, iden- tified as "basiorbital furrow" in Cyamodus kuhn- schnyderi by Nosotti and Pinna (1996).
The frontals are paired in Placochelys, as in all other cyamodontoids. Short anterolateral process- es of the frontal are embraced between the pre- frontals and nasals. These anterolateral processes of the frontal are shorter (i.e., less well developed) in Placochelys than in some other cyamodontoids. The concave lateral margin of the frontal broadly enters the dorsal margin of the orbit between the prefrontal and the postfrontal. A massive break running obliquely through the skull table and the right postorbital arch obscures sutural details of the frontoparietal suture. However, a posterolat- eral lappet of the frontal is clearly identifiable on the left side of the skull, indicating that the fron- toparietal suture was located at a level between the posterior margin of the orbit and the anterior margin of the temporal arch (Fig. 3A).
The postfrontal is a broad element that defines the posterodorsal margin of the orbit. Its ventral process tapers off along the posterior margin of the orbit but remains separated from the jugal by the postorbital. The posterior process is rather broad and extends backward to about the level of the anterior margin of the temporal fossa; it re- mains narrowly excluded from the anteromedial margin of the upper temporal fossa by a contact of the postorbital with the parietal. Posteriorly the postfrontal meets the parietal in an interdigitating suture, which slightly trends in an anteromedial direction. The medial margin of the postfrontal is more or less straight. The posterolateral margin of the postfrontal is deeply concave and angled in Placochelys, as it also is in Cyamodus kuhnschny- deri (Nosotti & Pinna, 1996), but unlike in Cy- amodus rostratus (Kuhn-Schnyder, 1965a), Pro- tenodontosaurus, or Psephoderma, where this bone has a less concave and more evenly curved posterolateral margin.
The unpaired (fused) parietal forms a rather rectangular and flat skull table characterized by extensive dermal ornamentation (Fig. 3A). The parietal skull roof carries four distinct tubercular protuberances. Posterolateral processes of the pa- rietal define the posteromedial margins of the up- per temporal fenestrae as well as the deeply con- cave occiput and meet the squamosal in an inter- digitating suture that runs across an osteodermal encrustation located at the posteromedial margin of the upper temporal fossa (left side of skull in Fig. 3A). The lateral margin of the skull table nar- rowly projects laterally beyond the descending
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
Fig. 3. Skull of Placochelys placodonta Jaekel (holotype, fafi Ob/2323/Vt.3): A, dorsal view; B, ventral view. Scale bar = 20 mm. For abbreviations, see p. 3.
flange of the parietal, which participates in the formation of a secondary lateral wall of the brain- case in a manner described in more detail below. The presence and position of the pineal fora- men in Placochelys remain uncertain. Jaekel (1907, PI. I) figured a relatively small pineal fo-
ramen located between the parietals at the level of the anterior margin of the upper temporal fossa but clearly behind the frontoparietal suture. Jae- kel's (1907, PI. I) drawing includes an element of reconstruction, however, because a massive break passes through the skull at precisely this level.
FIELDIANA: GEOLOGY
Fig. 3. Continued.
Huene (1931, PL I) nevertheless followed Jaekel's (1907) lead but increased the size of the pineal foramen located in the same position (i.e., behind the frontoparietal suture). Other cyamodontoids (Cyamodus, Protenodontosaurus, Psephoderma: see below) all have an equally large and anteriorly placed pineal foramen, but the frontal reaches far- ther back and narrowly approaches, or even en-
ters, its anterior margin. Although Huene's (1931) reconstruction appears plausible by comparison with other cyamodontoids, the cast of the skull of the holotype before its preparation by Strunz shows a distinct splint of bone embedded in the break in exactly the location where Jaekel (1907) and Huene (1931) placed the pineal foramen. The two possible conclusions are that this splint of
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
V.0 sq
B
Fig. 4. Skull of Placochelys placodonta Jaekel (holotype, fafi Ob/2323/Vt.3): A, left lateral view; B, occipital view. Scale bar = 20 mm. For abbreviations, see p. 3.
bone is in a natural position, in which case Pla- cochelys would lack a pineal foramen, or alter- natively, the splint of bone is in a displaced po- sition, in which case Jaekel's (1907) and Huene's (1931) reconstructions might be correct. The skull roof of mb.r. 1765 is severely damaged, with its anterior part missing. The break through the skull table lies at a level of the anterior margin of the upper temporal fossa. This again might indicate
an area of relative weakness, perhaps caused by a large and anteriorly placed pineal foramen.
The postorbital broadly enters the posteroven- tral margin of the orbit, from where it extends posteriorly to define the anterior and the greater part of the lateral margin of the upper temporal fossa. Huene (1931) believed the postorbital to extend far back along the medial margin of the upper temporal fossa, exposed in dorsal view lat-
10
FIELDIANA: GEOLOGY
eral to the parietal. This configuration of the post- orbital could not be confirmed for the holotype, where the postorbital meets the parietal at the an- teromedial margin of the upper temporal fossa, or for mb.r.1765. In lateral view, however, the post- orbital can be observed to form a posterior ver- tical flange that extends backward for a consid- erable distance below the overhanging rim of the parietal skull table (Fig. 4A), overlapping the ven- tral flange of the parietal and completing the an- terodorsal part of the secondary lateral wall of the braincase dorsal to the epipterygoid. At the pos- terodorsal corner of the orbit, the postorbital forms a stout ventromedial process that abuts the lateral surface of the anterodorsal corner of the epipterygoid. The posterolateral process of the postorbital extends along the lateral margin of the upper temporal fossa to a level behind the mid- point of the longitudinal diameter of the latter. This postorbital process does not narrow signifi- cantly at its posterior extremity, and meets the squamosal in an interdigitating suture (but see dis- cussion below).
The jugal carries a narrow anterior process that forms most of the ventral margin of the orbit dor- sal and medial to the maxilla (Figs. 3A, 4A). At the anterolateral margin of the subtemporal fossa, the jugal forms a distinct posteroventral lappet, ornamented with a pattern of radiating grooves and ridges. Behind the orbit, up to a level of about the midpoint of the postorbital arch, the jugal is sutured to the postorbital. More posteriorly, how- ever, the jugal separates from the postorbital and narrows to a pointed tip located somewhat in front of the posterior end of the postorbital, at about the level of the midpoint of the longitudinal diameter of the upper temporal fossa and about the lower third of the height of the temporal arch. Together, the dorsal margin of the jugal and the ventral mar- gin of the posterolateral process of the postorbital define a distinct V-shaped sutural pattern (the tip of the V pointing forward), thus embracing the anterior process of a bone whose identity remains to be discussed (see discussion below and Fig. 4A).
The squamosal of cyamodontoid placodonts is a very complex bone. Its body defines the pos- terolateral margin of the upper temporal fossa as well as the dorsal, lateral, and ventral margins of the posttemporal fossa. A dorsomedial process of the squamosal meets the posterolateral process of the parietal at the posteromedial margin of the up- per temporal fossa anteriorly and at the dorsal margin of the posttemporal fossa ventrally. These
sutural relations are very distinct, both in the ho- lotype of Placochelys and in mb.r. 1765 (Fig. 5A). In dorsal view, the contact of the squamosal with the parietal is bridged by an elongated der- mal encrustation located at the posterior margin of the upper temporal fossa (left side of the skull in Fig. 3A). This dermal encrustation tends to ob- scure the squamosal-parietal suture in the holo- type, particularly on the right side of the skull, whereas mb.r. 1765 clearly shows this interdigi- tating suture traversing the posterolateral part of the dermal encrustation (Fig. 5A). In lateral view and inside the temporal fossa, a narrow process of the squamosal can be followed along the dorsal margin of the posttemporal fossa, meeting the posterodorsal process of the epipterygoid in the anterodorsal corner of the posttemporal fossa. A similar contact of the epipterygoid with the squa- mosal at the anterodorsal corner of the posttem- poral fossa is also observed in Cyamodus rostra- tus (Kuhn-Schnyder, 1965a), but this character re- mains unknown for Cyamodus kuhnschnyderi (Nosotti & Pinna, 1996), Protenodontosaurus, and Psephoderma.
A ventromedial process of the squamosal curves around the lateral margin of the posttem- poral fossa, thereby establishing a broad ventro- lateral contact with the quadrate. Specimen mb.r. 1765 shows particularly well how this process of the squamosal extends anteromedially along the anterior aspect of the paroccipital process (Fig. 5B). Sutured to the opisthotic, this process of the squamosal forms the posterolateral margin of the pteroccipital foramen (sensu Nosotti & Pinna, 1993b, to be discussed in detail below), which is bordered anteromedially by the opisthotic.
A neomorphic process of the squamosal, the otic process sensu Nosotti and Pinna ( 1 993b), ex- tends anterolateral^ from the lower margin of the posttemporal fossa along the dorsal margin to the dorsomedial flange of the quadrate ramus of the pterygoid and meets the prootic at the anterolat- eral corner of the pteroccipital foramen. As a re- sult, the otic process of the squamosal forms the lateral margin of the pteroccipital foramen, which is bordered anterolaterally by the prootic (Figs. 3A, 6A, 6B).
It is the anterior and lateral relations of the squamosal that remain the most controversial as- pect of the dermatocranium in placodonts (Pinna, 1989; Zanon, 1989). The debate largely results from the difficulty of separating the squamosal from the quadratojugal within the temporal arch of placodonts. Pinna (1989; see also Pinna & No-
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
11
Fig. 5. Skull of Placochelys placodonta Jaekel (paratype, mb.r. 1765): A, dorsal view; B, ventral view. Scale bar 20 mm. For abbreviations, see p. 3.
sotti, 1989; Nosotti & Pinna, 1993b) reconstructed a very narrow squamosal that is barely exposed in the lateral view of the temporal arch both in Placodus and cyamodontoids. This leaves room for an expansive quadratojugal that broadly enters the lateral margin of the upper temporal fossa be- tween the postorbital and squamosal. Examination of all Placodus skulls in public repositories (Riep-
pel, 1995a) did not yield conclusive evidence to support Pinna's ( 1 989) reconstruction of the squa- mosal and quadratojugal in that taxon. In fact, none of the specimens shows a distinct and un- equivocal suture line separating the squamosal from the quadratojugal within the temporal arch (Rieppel, 1995a). The delineation of quadratoju- gal and squamosal in cyamodontoids is further
12
FIELDIANA: GEOLOGY
Fig. 5. Continued.
complicated because the lateral surface of the pos- terior part of the temporal arch is subject to the encrustation of dermal tubercles that obscure su- tural patterns. The only specimen apparently sup- porting Pinna's (1989) reconstruction of a small squamosal in cyamodontoids is Cyamodus cf. ros- tratus smns 17403 (Nosotti & Pinna, 1993b, Fig. 3), which shows what looks like a V-shaped con- tact of the squamosal with the quadratojugal at the posterolateral corner of the upper temporal fossa
(but see the detailed description of the specimen below).
A quadratojugal is unquestionably present in Placochelys, and it covers the lateral surface of the quadrate as in all other cyamodontoids. The suture separating the quadratojugal from the shaft of the quadrate is distinct in the occipital view of the skull (Fig. 4B). Equally distinct is the suture between quadratojugal and squamosal on the pos- terolateral aspect of the skull, at the level of the
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
13
pq. re
Fig. 6. Placochelys placodonta Jaekel (holotype, fafi Ob/2323/Vt.3); right lateral braincase wall, partially recon- structed. Scale bar = 20 mm. For abbreviations, see p. 3.
dorsal head of the quadrate. More anteriorly and laterally, this suture disappears under a dermal tu- bercle. Squamosal and quadratojugal therefore re- main indistinct from one another in the posterior part of the lateral surface of the temporal arch of the holotype (Fig. 4A). The medial surface of the temporal arch likewise offers no further clues.
In mb.r. 1765, the same sutural relations be- tween postorbital and jugal can be observed as were described for the holotype. The two bones again embrace a V-shaped anterior process of a posterior element (Fig. 5A). The pointed posterior tip of the jugal again lies somewhat in front of the posterior end of the postorbital and above the
14
FIELDIANA: GEOLOGY
ventral margin of the temporal arch. In this spec- imen, however, a suture line can be observed to extend backward from the posterior tip of the ju- gal in a more or less horizontal direction until it disappears under dermal encrustation on the pos- terolateral aspect of the temporal bar. In addition, the right side of the skull of MB. r. 1765 rather clearly shows the anterior tip of the quadratojugal tapering off along the ventral margin of the lower temporal arch below the jugal, reaching far for- ward to a level in front of the anterior margin of the upper temporal fossa (Fig. 5A). A similar an- terior extension of the quadratojugal can be re- constructed for the left temporal arch of the ho- lotype (Fig. 4A). A horizontal suture extending backward from the posterior tip of the left jugal can also be identified in "Macroplacus" raeticus (Schubert-Klempnauer, 1975; Rieppel, 1995a, Fig. 22), and, as in Placochelys, it appears to hor- izontally subdivide the posterior part of the tem- poral arch into a dorsal squamosal, broadly enter- ing the upper temporal fossa, and a ventral quad- ratojugal. This, in any case, is the most plausible reconstruction of the relationship of these two bones when observations on the holotype of Pla- cochelys placodonta and mb.r. 1765 are com- bined. All that is required is to link the suture that runs backward from the posterior tip of the jugal with the suture that separates the quadratojugal from the squamosal at the posterodorsal corner of the skull, a connection that is obscured by dermal encrustation of the temporal region of the skull. The anterior extent of the quadratojugal may be autapomorphic for Placochelys, but this character remains insufficiently known in other cyamodon- toids for meaningful comparison.
If the quadratojugal is reconstructed, as argued above, to form most of the ventral margin of the temporal arch and to be separated horizontally from the squamosal, the jugal and postorbital are left to embrace between themselves the anterior tip of the squamosal. In Placochelys, the anterior tip of the squamosal also reaches to a level in front of the anterior margin of the temporal fossa, an autapomorphy of the genus in comparison to other cyamodontoids.
In ventral view the skull shows the paired in- ternal nares to be separated by paired vomers (Fig. 3B). The sutural contact of the vomer with the maxilla is distinct at the anterior margin of the internal naris on both sides. The ventral surface of the rostrum is painted with resin, obscuring fur- ther details of the relationships of the vomer with the maxilla and premaxilla. However, there is no
indication of large anterior processes of the vo- mers, entering deeply between the premaxillaries, as is indicated in the figure published by Huene (1931). Posteriorly, the vomer contacts the pala- tine at the posteromedial corner of the internal naris.
As in all placodonts, the palatine is enlarged at the expense of the pterygoid. In Placochelys, it carries a smaller anterior and a much enlarged posterior tooth plate (Table 1 ). The palatine forms most of the posterior and lateral margin of the internal naris. Anteriorly and laterally, the palatine contacts the maxilla. At the anterior margin of the subtemporal fossa, the palatine carries a distinct lateral process that embraces the posterior end of the maxilla and contacts the jugal laterally (Fig. 3B). This lateral process cannot represent the ec- topterygoid, as indicated in the figure published by Huene (1931), because it carries on its dorsal surface the anterior tip of a groove that housed the palatoquadrate cartilage in the living animal. Posteriorly, the palatine meets the pterygoid in an interdigitating suture that curves around the pos- terior margin of the posterior palatine tooth plate. Within a distinct depression on the palatine-pter- ygoid suture lies the large, transversely oriented dental lamina foramen, located posteromedial ly to the posterior palatine tooth plate (Fig. 3B). A small dental lamina foramen is located postero- medially to the anterior palatine tooth plate in mb.r. 1765 (Fig. 5B), but a similar foramen is not distinct in the holotype. In both specimens, irreg- ular suture lines are observed at the palatine-pter- ygoid contact, delineating a triangular area in the midline of the palate between the two bones (Figs. 3B, 5B). It appears possible that the dermal palate includes a small heterotopic ossification between palatines and pterygoids. In lateral view, the pal- atine can be seen to extend backward medial to the pterygoid to meet the quadrate along the lat- eral margin of the palatoquadrate cartilage recess (described in more detail below); more anteriorly, the dorsal surface of the palatine carries a groove that housed the palatine process of a cartilaginous palatoquadrate, which persisted in the adult.
The pterygoids complete the posterior part of the dermal palate. They meet one another in an interdigitating suture, which in the holotype shows an irregularly curved course (Fig. 3B). Lat- erally, the pterygoid forms a prominent, longitu- dinally oriented ventral flange and extends ante- riorly to the level of the posterior third of the lon- gitudinal diameter of the posterior palatine tooth plate. The ventral pterygoid flange forms a single
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
15
ventral projection. Compared to other cyamodon- toids (see below), the palatal exposure of the pter- ygoid is relatively long in Placochelys: dividing the distance from the posterior margin of the pter- ygoid to the posterior dental lamina foramen by the distance from the posterior dental lamina fo- ramen to the posterior margin of the internal naris yields a quotient of 0.45. Posterolateral^, the pterygoid forms a short quadrate ramus that in ventral view meets the anteromedial aspect of the quadrate in an interdigitating suture. The quadrate ramus of the pterygoid carries a broad dorsal flange that extensively overlaps a broad antero- medial flange of the quadrate, thus obliterating the cranioquadrate passage anteriorly. The overlap of these two bones is well exposed in lateral view, as will be discussed in more detail below.
The presence or absence of an ectopterygoid in cyamodontoid placodonts again remains a matter of debate. Although generally assumed to be pre- sent (Huene, 1931; Pinna & Nosotti, 1989; No- sotti & Pinna, 1996), its delineation from the pal- atine and pterygoid has been notoriously difficult. The presence of an ectopterygoid in Psephoderma could not be ascertained (personal observation), and an ectopterygoid is positively absent in Pro- tenodontosaurus (Nosotti & Pinna, 1998; personal observation). The lateral view of the skull (holo- type) of Placochelys shows the pterygoid to be sutured to the quadrate posterodorsally and to the palatine anterodorsally (Figs. 6A, 6B). As is also seen in ventral view, the anterolateral tip of the pterygoid reaches to about the posterior third of the longitudinal diameter of the posterior palatine tooth plate. Exposed in lateral view in front of the pterygoid, a suture line appears to separate from the palatine a small, splintlike element that might represent an ectopterygoid. If so, the ectoptery- goid would only line the anteromedial margin of the subtemporal fossa with hardly any ventral ex- posure at all. However, the supposed palatine-ec- topterygoid suture seen in lateral view could also represent a crack, since no clear separation of an ectopterygoid from the palatine can be seen in ventral view.
Two elements ossify in the palatoquadrate of reptiles, the quadrate and the epipterygoid. The quadrate of Placochelys can be described as being composed of a shaft and a broad anteromedial wing. The posterior aspect of the quadrate shaft is distinctly concave. The lateral surface of the shaft is covered by the quadratojugal (Fig. 3A). The mandibular condyle of the quadrate is bipar- tite, a central concavity matching the saddle-
shaped articular surface of the lower jaw. A shal- low stapedial recess is located on the anteromedial aspect of the quadrate shaft, located narrowly above the mandibular condyle (Figs. 3B, 5B); it must have received the (cartilaginous?) distal end of the stapes. In the holotype, a relatively large foramen can be identified, located lateral to the dorsal head of the quadrate, between the latter and the squamosal (Fig. 4B). The quadratojugal is ex- cluded from this foramen on the left side but nar- rowly enters its ventral margin on the right side. The foramen may have served the passage of a lateral branch of the internal carotid or stapedial artery respectively to the temporal musculature. A comparable foramen is not distinct in mb.r. 1765 (obscured by deformation of the skull?) and is ab- sent in all other cyamodontoids (but see the dis- cussion of Macroplacus, below). The broad an- teromedial wing of the quadrate forms the sloping posterior wall of the temporal fossa from which the posterior part of the external jaw adductor must have originated. It is well exposed in lateral view, as it demarcates the posterolateral margin of a deep palatoquadrate cartilage recess (Fig. 6). The broad and complex epipterygoid is the dominant element in the secondary lateral wall of the braincase. The bone can be described as con- sisting of two parts, an anterior portion with a deeply concave lateral surface and a posterior por- tion with a distinctly convex lateral surface (Fig. 6). These two parts of the epipterygoid may cor- respond to the anterior "palatal ramus" and pos- terior "quadrate ramus" of the epipterygoid of Cyamodus kuhnschnyderi (Nosotti & Pinna, 1996), although the opening that separates these two parts of the epipterygoid in the latter taxon is absent in Placochelys (the lateral opening within the epipterygoid of Cyamodus kuhnschnyderi ap- pears to be the result of incomplete ossification; Nosotti & Pinna, 1996: 27). The posterior part of the epipterygoid of Placochelys shows a deeply concave posterior margin that defines the anterior margin of the trigeminal incisure, enclosed be- tween the epipterygoid and the prootic. Postero- dorsally, the epipterygoid is extended into a long, slender process that runs across the dorsal margin of the prootic and meets the squamosal in the an- terodorsal corner of the posttemporal fossa (Fig. 6). Posteroventrally, the epipterygoid narrowly contacts the prootic on the left side of the skull (of the holotype), thus closing the trigeminal in- cisure ventrally; a similar contact is absent on the right side of the skull. The ventral margin of the posterior part of the epipterygoid shows a surface
16
FIELDIANA: GEOLOGY
of unfinished bone, which overhangs the medial margin of the palatoquadrate cartilage recess. The anterior part of the epipterygoid is sutured to the dorsal surface of the palatine and reaches far for- ward to a level dorsal to the posterior palatine tooth plate (Figs. 5A, 5B). The anteroventral tips of the epipterygoids of both sides converge to- ward the midline of the skull.
The prootic is exposed in lateral view (Fig. 6) as it emerges from below the narrow posterodor- sal process of the epipterygoid. Its lateral surface is distinctly convex. The anterior margin of the prootic defines the posterior margin of the trigem- inal incisure. The ventral margin of the prootic broadly meets the neomorphic otic process of the squamosal in an oblique suture. These two ele- ments together form the ventral margin of the posttemporal fossa, which also represents the an- terior margin of the pteroccipital foramen (Figs. 3A, 6A). Laterally, the otic process of the squa- mosal together with the ventral margin of the pro- otic define the posteromedial margin of the pala- toquadrate cartilage recess.
The palatoquadrate cartilage recess is a char- acter of all cyamodontoid placodonts included in this study whose skulls are adequately preserved and prepared to reveal the morphology of the lat- eral braincase wall (Figs. 3A, 4A, 6). It is repre- sented by a characteristic cleft with a triangular outline of its posterior part, bordered laterally by the medial wing of the quadrate and by the pala- tine, bordered medially by the squamosal, prootic, and epipterygoid, and floored by the pterygoid posteriorly and by the palatine anteriorly. This re- cess continues anteriorly as a distinct groove on the dorsal surface of the palatine, which in its an- terior part turns laterally toward the anterior cor- ner of the subtemporal fossa. In the living animal, this recess and groove must have housed cartilage of the palatoquadrate, which persisted in the adult in its classic position dorsal to the dermal palate (pterygoid and palatine) and which connected the two ossifications within the palatoquadrate (i.e., the quadrate and epipterygoid).
The pteroccipital foramen (Nosotti & Pinna, 1993b) is another synapomorphy shared by all cy- amodontoids included in this study with adequate- ly preserved and prepared skulls. Whereas Pla- codus retains a complete cranioquadrate passage (Rieppel, 1995a), the latter is obliterated in cy- amodontoids in its anterior part by fusion of the dermal palate to the basicranium and by the broad overlap of the dorsal wing of the pterygoid with the anteromedial wing of the quadrate. This over-
lap also closes a gap which in the skull of Pla- codus persists between the quadrate and pterygoid laterally and the lateral braincase wall medially, and through which the stapedial (temporal) artery reached the jaw adductor musculature. In cyamo- dontoid placodonts, the stapedial (temporal) ar- tery reaches the jaw adductor musculature through the pteroccipital foramen, located at the ventral margin of the posttemporal fossa (Figs. 3, 5, 6). It is bordered anteriorly by the otic process of the squamosal and the prootic, and posteriorly by the squamosal and the opisthotic, the latter ex- tended into the paroccipital process. As the sta- pedial (temporal) artery branches off the internal carotid within the posterior part of the crani- oquadrate passage, it turns dorsally to pass through the pteroccipital foramen in front of the paroccipital process; it reemerges from the same foramen at the ventral margin of the posttemporal fossa, through which it gains access to the tem- poral musculature.
The occipital view of the skull features the rel- atively large posttemporal fossae, bordered dor- sally by the occipital flange of the parietal and squamosal and ventrally by the slender paroccip- ital process (Fig. 4B). The supraoccipital is a trap- ezoidal element carrying a low median crest. It is obliquely inclined, contacts the parietal in an in- terdigitating suture anteriorly, and defines the dor- sal margin of the foramen magnum posteriorly. The lateral margin of the foramen magnum is formed by the exoccipital. The exoccipitals of ei- ther side do not meet dorsal to the occipital con- dyle, which is formed by the basioccipital only. The occipital condyle of mb.r. 1 765 shows a dis- tinct notochordal pit on its posterior surface (Fig. 5B). The metotic (jugular or vagus) foramen (Rieppel, 1985) is enclosed between the exoccip- ital and the opisthotic. Neither the holotype nor mb.r. 1765 shows an internal subdivision of the metotic foramen that would have separated the exit of the roots of the glossopharyngeal, vagus, accessory, and hypoglossal nerves. Below the me- totic foramen and lateral to the occipital condyle, the braincase forms two ventral processes on ei- ther side of the skull. The medial one, close to the occipital condyle, is the ventrally directed ba- sioccipital tuber. A ventral process of the opis- thotic is located lateral to the basioccipital tuber. The two ventral processes together enclose an opening or foramen, which may have trapped the internal carotid at the entry into the cranioquad- rate passage. In the holotype, the basioccipital tu- ber and the opisthotic process do not extend ven-
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
17
|
Hfr |
|||
|
"■M |
|||
|
. §5 ' |
|||
|
'''•'•'■ |
|||
|
■ ^H |
'''': ^« |
||
|
■ ^ ■■ ._ |
|||
|
SS^m''-y^: |
|||
|
^K^> , |
|||
|
iHbP^Pj^B |
|||
|
t,lBRn'i!''!-y- ■■» v |
•'--.""'' • : |
||
|
Ik^Se^- ■ |
■\^B 5? |
||
|
W^Lsr <ii. |
■"•• v'::-\':>v V' |
||
|
Byjfc. ■"*>« |
|||
Fig. 7. Placochelys placodonta Jaekel (paratype, mb.r. 1765); contact between the left paroccipital process (opisthotic) and the left squamosal. Scale bar = 20 mm.
trally to meet the basisphenoid or the posterior margin of the pterygoid, and the two structures also fail to make a ventral contact with each other, such that the passage of the internal carotid be- tween the opisthotic and basioccipital is open ven- trally (Fig. 4B). In mb.r. 1765, the opisthotic pro- cess and the basioccipital tuber are more strongly developed and meet each other ventral to the pas- sage of the internal carotid, which is thereby cap- tured in a closed foramen (Fig. 5B). Whether the close approximation of these two processes to the basisphenoid and pterygoid reflects natural rela- tions or results from dorsoventral compression of the skull is difficult to determine. But given the relatively larger size of both structures in the specimen mb.r. 1765, it appears possible that the area was damaged during preparation of the ho- lotype. In both skulls of Placochelys, the opis- thotic is pierced by a small foramen located just above the ventral flange. The function of this fo- ramen remains unknown; separate exit(s) for the root(s) of the hypoglossal nerve would be ex- pected to be located within the exoccipital medial to the passage of the internal carotid.
Laterally, the opisthotic extends into a slender paroccipital process. The distal end of the paroc- cipital process abuts a distinct buttress, located on the lower surface of the squamosal medial to the dorsal head of the quadrate (Figs. 3B, 5B, 7). In mb.r. 1765, but not in the holotype, the squamosal buttress can be seen to extend into a medially di- rected ridge on the lower surface of the squamo- sal, which follows the anterodorsal aspect of the paroccipital process and meets the opisthotic in an oblique suture at the posterior margin of the pteroccipital foramen (see above for a detailed de-
scription of the squamosal). Medial to this con- tact, mb.r. 1765 also displays the vestibular (oval) fenestra of the otic capsule, enclosed by the pro- otic anteriorly and the opisthotic posteriorly (Fig. 5B). Opisthotic and prootic remain separate, as is indicated by a distinct suture at the dorsal margin of the vestibular fenestra. The vestibular fenestra lies deep inside the posterior part of the cranio- quadrate passage. In front of it, the cranioquadrate passage is obliterated by the fusion of the dermal palate to the basicranium and by the broad overlap of a dorsal flange of the pterygoid with the an- teromedial flange of the quadrate. No stapes is preserved in mb.r. 1765, but an imaginary straight line that extends from the vestibular fenestra in a posterolateral direction connects the latter with the stapedial recess on the quadrate. The location of the pteroccipital foramen behind this imaginary line would seem to indicate that the stapedial ar- tery, which branches off from the internal carotid after the latter has entered the cranioquadrate pas- sage through the gap between the basioccipital tu- ber and the ventral opisthotic process, would have passed behind the stapes on its way to the tem- poral region of the skull. However, the location of the pteroccipital foramen with respect to the par- occipital process in the three-dimensionally pre- served holotype of Placochelys suggests that the stapedial artery passed in front of the stapes in- stead.
The left side of the occiput of the holotype shows what looks like a distally bifurcated par- occipital process, or a sturdy process emerging from behind and below the paroccipital process (Figs. 4B, 8). This element was tentatively iden- tified as the distal end of a massive stapes by No- sotti and Pinna (1996: 33). Following this inter- pretation, cyamodontoid placodonts would have lost an impedance matching middle ear and in- creased the volume of the stapes in a system that relied on sound transmission through bone only. Close inspection of the holotype shows, however, that the supposed stapes is not distinctly separated from the paroccipital process. Indeed, the two structures appear to be separated by breaks only, and where breaks are absent, the two structures appear to be confluent. In search of alternative explanations for the supposed stapes, reference may be made to the "lateral tubercle" on the par- occipital process of Cyamodus kuhnschnyderi: "By its postero ventral margin, close to the squa- mosal, the paroccipital process bears a small, downward and posteriorly projecting tubercle" (Nosotti & Pinna, 1996: 19). This structure, how-
FIELDIANA: GEOLOGY
Fig. 8. Placochelys placodonta Jaekel (holotype, fafi Ob/2323/Vt.3); the skull in occipital view, showing the left paroccipital process. For further discussion see text. Scale bar = 20 mm.
ever, is much smaller than the supposed stapes in the holotype of Placochelys, and a comparable "lateral tubercle" is absent on the perfectly pre- served paroccipital processes of mb.r.1765. In- stead, the latter specimen shows a slight expan- sion of the ventral surface of the paroccipital pro- cess into a weakly protruding flange with a rugose surface indicating muscle attachment, probably of the longissimus capitis muscle. Even if broken, this flange would not be large enough to match the supposed stapes in the holotype. However, mb.r. 1765 also shows a distinct ridge on the ven- tral surface of the squamosal, originating at the squamosal buttress for the paroccipital process and extending medially along the anterodorsal as- pect of the latter (Fig. 5B). A comparable ridge is not easily identified in the holotype, and perhaps the supposed stapes could represent a broken seg- ment of this ridge. In the end, however, the iden- tity of the bone fragment tentatively identified as stapes in the holotype of Placochelys remains de- batable, as was also noted by Nosotti and Pinna (1996). Should it indeed represent the stapes, it
would add further weight to the argument that the stapedial (temporal) artery passed in front of the latter on its way through the pteroccipital fora- men.
Breakage of the skull roof in mb.r. 1765 ex- poses details of the basicranium in dorsal view (Fig. 5A). Easily identified is the sella turcica on the basisphenoid-parasphenoid complex, located in front of a distinct dorsum sellae and pierced by two foramina through which the cerebral carotids must have gained access to the cranial cavity. The basisphenoid-parasphenoid complex is broken shortly in front of the sella turcica. The passage of the cerebral carotids through foramina in the sella turcica indicates that the internal carotid must have pursued an intracranial course after having traversed the posterior part of the cranio- quadrate passage and after having given rise to the stapedial (temporal) artery. Unfortunately, none of the available skulls of cyamodontoid pla- codonts allows detailed reconstruction of the course of the internal carotid canal through the basicranium. Only in Cyamodus kuhnschnyderi is there some indication that the internal carotid en- tered the basicranium between the basisphenoid and the otic capsule at the depth of the cranio- quadrate passage. The internal carotid likewise passes between the basisphenoid and the otic cap- sule in the eosauropterygian genera Nothosaurus and Simosaurus (Rieppel, 1994a). Continuing an- teriorly within this basicranial canal, the internal carotid must have bifurcated, the dorsal branch emerging through the sella turcica as cerebral ca- rotid and the anterior branch continuing within the basicranial canal as palatine artery. This recon- struction of the course of the internal carotid rais- es the question as to how the palatine artery gained access to the soft palate, which it supplies with arterial blood. The only foramina in the der- mal palate located at a level in front of the sella turcica are those identified above as dental lamina foramina. The large foramina located posterome- dial to the posterior palatine tooth plates lie in close proximity to the foramina for the cerebral carotids through the sella turcica, which indicate the level of bifurcation of the internal carotid. It is conceivable that the palatine artery could have gained access to the soft palate by passing through these posterior dental lamina foramina.
Similar questions relate to the passage of the lateral head vein and the profundus branch of the trigeminal nerve through the cavum epiptericum, which, in a generalized reptile skull, constitutes the anterior continuation of the cranioquadrate
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
19
passage, obliterated in cyamodontoids. mb.r. 1765 shows the basisphenoid-parasphenoid complex to lie dorsal to the dermal palatines (Fig. 5A). No ossified base of the pila antotica (clinoid process) can be seen ascending dorsally from the lateral margin of the basisphenoid at the level of the sella turcica or dorsum sellae. There appears to be no ossification in the primary lateral wall of the braincase in Placochelys, in contrast to Placodus, where such an ossification was described as an "alisphenoid bridge" by Broili (1912).
The epipterygoid is located well lateral to the outer margin of the basisphenoid-parasphenoid complex, the latter representing the level of the primary lateral wall of the braincase. The body of the epipterygoid is obliquely inclined with respect to the dorsal surface of the palatine, meeting the latter at an acute angle. Through this relation of the bones, a narrow space becomes enclosed be- tween the medial surface of the epipterygoid and the dorsal surface of the palatine, as can also be seen in a horizontally split skull fragment of a cyamodontoid from the Muschelkalk of Makhtesh Ramon, Negev (HUJ-Pal. 220; Rieppel, Mazin & Tchernov, 1999, Fig. 7). mb.r. 1765 is unusual, however, in that a vertical lamina of bone appears to descend from the margins of the broken skull roof to the dorsal surface of the palatine medial to the epipterygoid but lateral to the lateral margin of the basisphenoid-parasphenoid complex (Fig. 5A). This vertical lamina of bone appears to be separated from the epipterygoid by a gap that opens in a minute "foramen" or cleft close by the anterior tip of the epipterygoid on its medial side. Because it has no relationship to the lateral mar- gin of the basisphenoid-parasphenoid complex, this vertical lamina of bone cannot represent an ossification of the primary lateral wall of the braincase. Also, the anterior "foramen" located between the vertical lamina and the epipterygoid appears to be too small to represent the anterior opening of a cavum epiptericum enclosed be- tween the vertical lamina and the epipterygoid. It therefore appears that the epipterygoid has been longitudinally split because of dorsoventral com- pression of the skull, resulting in a vertical lamina that appears to be separated from the laterally placed part of the epipterygoid. Alternatively, it might be assumed that the epipterygoid broke along a horizontal line and that the dorsal part of the bone, which is very thin at its anterior margin in the holotype, was pushed ventrally medial to the base of the epipterygoid because of dorsoven- tral compression. This interpretation is supported
by the observation that neither the holotype of Placochelys nor the horizontally split skull of the cyamodontoid from Makhtesh Ramon (HUJ-Pal. 220) show a vertical lamina of bone separated from, and located medial to, the basal part of the epipterygoid that, because it cannot be an ossifi- cation of the primary lateral wall of the braincase, would have to represent a vertical downgrowth from the parietal, frontal, postorbital, or a com- bination thereof.
Whereas the maxillary and mandibular branch- es of the trigeminal nerve would have emerged from the trigeminal incisure behind the epiptery- goid, the profundus branch would have emerged anteriorly from the cavum epiptericum (i.e., from the gap between the epipterygoid and the lateral margin of the basisphenoid-parasphenoid com- plex). The course of the lateral head vein, which in reptiles passes through the cranioquadrate pas- sage and the cavum epiptericum, is less easily re- constructed for cyamodontoids because of the obliteration of the anterior part of the cranioquad- rate passage (Nosotti & Pinna, 1996). Kuhn- Schnyder (1960) suggested that the lateral head vein would have escaped the posterior part of the cranioquadrate passage through the pteroccipital foramen, rather than passing through the cavum epiptericum. This certainly remains a possibility, but it must be remembered that there is a certain plasticity in the differentiation of the cranial ve- nous system. During embryonic development, the primary head vein (vena capitis medialis) forms loops and sinuses surrounding the roots of the cra- nial nerves and the auditory sac as it becomes replaced, at least in part, by the lateral head vein (vena capitis lateralis), and as differences in the differentiation of these veins persists among rep- tiles in general, and in squamates in particular (van Gelderen 1924; Goodrich, 1930), the cranial veins of cyamodontoids may have been differen- tiated in a pattern that may not be easily compared with that seen in extant reptiles.
Another unsolved problem of the cranial anat- omy of cyamodontoid placodonts is the course of the palatine and hyomandibular branches of the facial nerve. In reptiles, the facial nerve usually exits through a foramen located between the com- missura basicapsularis and the commissura prae- facialis of the endocranium, which link the otic capsule with the basal plate and which both ossify as part of the prootic. However, the prootic of Placochelys lacks a foramen for the exit of the facial nerve (Fig. 6), as is also the case in Cy- amodus kuhnschnyderi (Nosotti & Pinna, 1996).
20
FIELDIANA: GEOLOGY
Table 2. Measurements of the dentary tooth plates of Placochelys placodonta (holotype, mafi Ob/2323/ Vt.3). All measurements in mm; approximate values in parentheses.
|
left |
right |
|||
|
longi- :uJnul0 |
trans- verse 0 |
longi- tudinal 0 |
trans- vcrsc 0 |
|
|
anterior dentary tooth |
19.1 |
12.0 |
18.0 |
12.0 |
|
posterior dentary tooth |
29.0 |
20.3 |
28.5 |
(21) |
One specimen (mfsn 1923GP) of Protenodonto- saurus was observed to show an internal subdi- vision of the trigeminal incisure by a vertical strut of bone (Nosotti & Pinna, 1996). Should the pos- terior division of the incisure have served the exit of the hyomandibular branch of the facial nerve, the latter would have to have reached the middle ear region through the pteroccipital foramen (No- sotti & Pinna, 1996). Similarly, the paratype of Cyamodus kuhnschnyderi (smns 1 6270) shows, on the right side of the skull, a foramen at the ventral margin of the prootic closely behind the trigemi- nal incisure. Should this foramen have served as the exit of the hyomandibular branch of the facial nerve, it would again have to be assumed that the latter reached the middle ear cavity through the pteroccipital foramen. Alternatively, the hyoman- dibular branch might have exited the prootic in close proximity to the vestibular (oval) fenestra deep in the cranioquadrate passage, a region not well exposed in any of the cyamodontoid skulls available for study. The palatine branch of the fa- cial nerve, on the other side, appears to have pur- sued an intracranial course, joining the internal carotid and palatine artery respectively on their way through the basicranium (Nosotti & Pinna, 1996).
Morphological Description of the Lower Jaw
The lower jaw (Fig. 9) was separated from the skull by Strunz after the description of the holo- type by Jaekel (1907). As is true for the rostrum, the tip of the mandibular symphysis is incomplete. The retroarticular processes also appear to be in- complete, probably because of damage inflicted on them through Strunz's removal of dermal tu- bercles that were fused to the articular. The tip of the right coronoid process is incomplete, whereas the tip of the left coronoid process, although sub- ject to some damage, appears to retain its original
height at least at its anterodorsal corner. The total length of the (better preserved) left mandible is 122 mm; its total height at the anterodorsal mar- gin of the coronoid process is 56.5 mm. Each den- tary carries on its posterior part two tooth plates, of which the posterior one is distinctly larger than the anterior one (Table 2).
The dentaries form the deep, massive, and elon- gated mandibular symphysis, with some partici- pation of the splenials at the ventral margin of the latter. Posteriorly, the dentary extends to a level well behind the anterior margin of the large cor- onoid process, thus carrying the large posterior tooth plate to a position partially medial to the anterior part of the coronoid process. The dentary is broadly exposed in lateral view, whereas the splenial gains only a limited lateral exposure along the ventral margin of the anterior part of the mandible. These relations are reversed on the medial aspect of the lower jaw, with a relatively narrow exposure of the dentary and a broad ex- posure of the splenial. The two bones are sepa- rated from one another by the prearticular.
The coronoid process is formed exclusively by the large coronoid, the dominant element on the lateral surface of the lower jaw. As is typical for most cyamodontoids, the coronoid reaches far down, closely approaching the ventral margin of the lower jaw (Fig. 9A). An anteroventral process of the coronoid, which overlaps the posterior part of the dentary, tapers to a pointed tip at a level below the anterior dentary tooth plate. The pos- teroventral margin of the coronoid is rounded in- stead (distinct on the right mandible only). The lateral surface of the coronoid is turned outward along its ventral margin, forming a distinct shelf that overhangs the contact of the splenial with the angular along the ventral margin of the lower jaw. This shelf delineates the ventral extent of the area of insertion for superficial jaw adductor muscle fibers on the lateral surface of the coronoid.
The posterior part of the lower jaw includes the large angular, which extends anteriorly to a level in front of the apex of the coronoid process, where it tapers off in an overlapping sutural contact with the splenial. Behind the coronoid process, the lat- eral exposure of the angular is larger (deeper) than that of the surangular. The suture between the sur- angular and angular is not very distinct, however, but seems to correspond to a distinct and curved ridge on the lateral surface of the lower jaw. Con- cave ventrally, this ridge delineates the insertional facet for the superficial portion of the pterygoi- deus muscle. Along the dorsolateral edge of the
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
21
Fig. 9. Lower jaw of Placochelys placodonta Jaekel (holotype, fafi Ob/2323/Vt.3): A, lateral view; B, medial view. Scale bar = 20 mm. For abbreviations, see p. 3.
mandible and behind the coronoid process, the surangular forms a distinct, dorsally protruding rim lateral to the articular surface of the lower jaw, but does not appear to participate in the for- mation of this articular surface itself. The articular surface of the mandibular joint is saddle-shaped, matching the surface of the mandibular condyle of the quadrate. The articular extends posteriorly into a short retroarticular process. The retroarti- cular process is very distinct in the placodontoid genera Placodus and Paraplacodus (Rieppel, 1995a). In the better known genus Placodus, the retroarticular process is long, slender, and turned slightly upward. In contrast, the retroarticular pro- cess of cyamodontoid placodonts is generally
short, stout, and has a posteroventrally sloping surface, as is seen in a perfectly preserved left mandible (Fig. 10) of Cyamodus hildegardis (pi- muz T2796; Kuhn-Schnyder, 1959, PI. 1, Fig. b; Pinna, 1992, PI. 10, Fig. 8). In Placochelys the chorda tympani foramen is located on the dorsal surface of the retroarticular process closely behind the articular facet within the articular.
The prearticular is a large element that, together with the angular, closes the large Meckelian canal medially (Fig. 9B). The latter opens posteriorly on the medial aspect of the lower jaw, where it forms a deep adductor fossa. The adductor fossa has a considerable longitudinal extension medial to the coronoid process and extends downward to
22
FIELDIANA: GEOLOGY
ang
Fig. 10. Lower jaw of Cyamodus hildegardis Peyer (pimuz T2796) in lateral view. Scale bar = 20 mm. For abbreviations, see p. 3.
the ventral margin of the lower jaw. Two parallel, longitudinally oriented ridges are located on the dorsal surface of the angular at the bottom of the adductor fossa. These may have secured the at- tachment of tendinous plates of the bodenaponeu- rosis, into which inserted the fibers of the medial and deep portions of the external jaw adductor muscle. Similar ridges are seen at the bottom of the adductor fossa in Placodus (Rieppel, 1995a).
Dermal Ornamentation of the Skull
Cyamodontoid placodonts generally show a tu- bercular dermal ornamentation of the temporal bones of the skull. A closer look at Placochelys reveals two different components in the dermal ornamentation of the skull. The postfrontal, post- orbital, and the parietal in particular show a pat- tern of ornamentation that might be called dermal encrustation, resulting in the formation of projec- tions or bosses (Figs. 3A, 4A). The ossification center of the postfrontal and postorbital is elevat- ed into a low and blunt apex, from which grooves and ridges radiate toward the margins of the bone. Reflecting the growth pattern of the underlying bone, this dermal encrustation appears to have de- veloped simultaneously with the ossification of the underlying bone itself. The same appears to apply to the four relatively low and ill-defined tu- bercular encrustations observed on the parietal skull table and to the encrustations across the pa- rietal-squamosal suture at the posteromedial mar- gin of the upper temporal fossa. Whether these projections or bosses reflect areas of epidermal scales remains unknown, although it is likely.
A much more pronounced pattern of dermal or- namentation is found on the surface of the pos- terior part of the squamosal and quadrat ojugal in
the form of distinct dermal tubercles that are par- ticularly well-developed at the posterolateral mar- gins of the temporal region of the skull (Figs. 3, 4). These tubercles have a sharply defined base, which appears to have secondarily fused to the underlying bone. Their position correlates in no way with the growth pattern of the underlying bone. In specimens of Cyamodus hildegardis (pi- muz T4763, T4771; Peyer, 1931a), it can be seen that the large temporal tubercles located on the posterior margin of the squamosal (pimuz T4771; Pinna, 1992, Fig. 7, right side of skull) match those aligned along the lateral margins of the car- apace (pimuz, T4763; Peyer, 1931a, PI. 15, right posterolateral margin of dorsal shield) both in size and shape. To judge from the mode of their at- tachment to the surface of the underlying tempo- ral bones of the skull, it appears that these large tubercles initially developed independently from the underlying bone and only later fused to the surface of the latter.
The inference that these large and distinct der- mal tubercles are secondarily fused to the under- lying bone is supported by Jaekel's (1907) figure of the holotype of Placochelys, which shows on both sides of the skull two such tubercles fused to the dorsal surface of the retroarticular process closely behind the mandibular joint, formed by the articular. As the latter represents an endoskel- etal element, dermal tubercles must have fused to it after its ossification. The dermal tubercles on the retroarticular process were removed through preparation by Strunz.
Comparison of the Cranial Anatomy of Placochelys placodonta with That of Other Cyamodontoid Placodonts
A number of well-preserved and well-prepared skulls of cyamodontoid placodonts are available that allow a detailed comparison with the skull of Placochelys placodonta. These include Cyamodus rostratus (Miinster, 1839) (umo BT 748; original of Drevermann, 1928, and Kuhn-Schnyder, 1965a; smns 17403, referred specimen); Cyamo- dus kuhnschnyderi Nosotti and Pinna, 1993a (smns 15855, smns 16270; see also Nosotti & Pin- na, 1996); Henodus chelyops v. Huene, 1936 (gpit "specimens I and II," syntypes of Huene, 1936; "specimens IV and VI," collected in 1959). Ma- croplacus raeticus Schubert-Klempnauer, 1975 (bsp 1967 I 324); Protenodontosaurus italicus
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
23
Fig. 11. Skull of Cyamodus rostratus Miinster (holotype, umo BT 748): A, dorsal view; B, ventral view; C, left lateral view. Scale bar = 20 mm.
Pinna, 1990b (mfsn 1819GP, mfsn 1923GP; see also Nosotti & Pinna, 1998), and Psephoderma alpinum Meyer, 1858a, b (msnm V471, see also Pinna & Nosotti, 1989). All of the taxa mentioned above have been subject to personal observation and form the basis of the following comparative analysis.
The Cranial Anatomy of Cyamodus rostratus (Munster, 1839)
Cyamodus rostratus is the type species of the genus Cyamodus (H. v. Meyer, 1863). It is rep- resented by an incomplete skull (umo BT 748, ho-
lotype) from the lower upper Muschelkalk (mol) of Bayreuth (Bindlach, Lainecker Hohenzug), Ba- varia, Germany (Fig. 11). umo BT 2172 and smf R-4040 are two isolated lower jaws from the same locality, smns 17403 is an incomplete skull from the lower upper Muschelkalk (Trochitenkalk, mol) from the Burrer Quarry, Gaismuhle near Crailsheim, southern Germany.
Cyamodus rostratus may represent the most generalized cyamodontoid with respect to many features of its cranial anatomy (Figs. 12, 13). The skull of Cyamodus rostratus (umo BT 748) is in- completely preserved and was redescribed by Kuhn-Schnyder (1965a). Both temporal arches are missing, as is the right suspensorium. The basi-
24
FIELDIANA: GEOLOGY
cranial length (tip of rostrum to basioccipital con- dyle) measures 106 mm. The longitudinal diam- eter of the left orbit measures 27.0 mm; the trans- verse (vertical) diameter of the same orbit mea- sures 25.0 mm. The longitudinal diameter of the left upper temporal fossa approximates 60 mm; its transverse diameter can be estimated to mea- sure 43 mm. By comparison with other cyamo- dontoids, the skull of Cyamodus rostratus shows a relatively high and narrow temporal region and deep orbits facing laterally.
The premaxillaries form a broad, short rostrum. Each premaxilla bears two teeth, which are bul- bous and rounded but still form an anterior trans- verse crest reminiscent of chisel-shaped anterior premaxillary teeth seen in the sister-taxon of cy- amodontoids, Placodus (Rieppel, 1995a). Short posterior (nasal) processes of the premaxillae en- ter between the external nares and meet the nasals in a V-shaped suture at about the level of the mid- point of the longitudinal diameter of the external nares (Fig. 12 A). In ventral view, distinct yet slen- der posterior (vomerine) processes of the premax- illae meet the equally slender vomers at a level in front of the midpoint of the longitudinal diameter of the internal naris (Fig. 12B). The margin of the internal naris is fairly complete on the right side of the skull and shows that the premaxilla remains excluded from the anterior margin of the external naris by the vomer and maxilla (the suture be- tween the latter two elements is indistinct). In lat- eral view, the transverse process of the premaxilla reaches backward to a level slightly in front of the posterior margin of the external naris. It forms the anteroventral margin of the external naris and meets the maxilla in a V-shaped suture (the apex pointing backward).
The nasals are broad, leaf-shaped structures (Fig. 12A). They define the posteromedial margin of the external naris and meet each other along the midline, thereby broadly separating the pre- maxilla from the frontal. Each nasal forms a short and tapering posterior process; together these em- brace an anteromedial process formed by the fron- tals. Posterolaterally, the nasal meets the prefron- tal, thus separating the frontal from the maxilla.
The frontals are paired and elongated elements that posteriorly reach to a level behind the anterior margin of the upper temporal fossa (Fig. 12A). Anterolaterally, each frontal forms a distinct yet slender anterolateral process that enters between the prefrontal and the nasal. Between the prefron- tal and the postfrontal, the frontal broadly enters the dorsal margin of the orbit. The orbital margin
of the frontal is distinctly concave. Although closely approaching the pineal foramen, the fron- tals remain excluded from it by the parietal. The posterior end of the frontal is rather broad and meets the parietal in a more or less transversely oriented, weakly interdigitating suture. Each fron- tal carries three distinct grooves on its posterior (postorbital) part that converge toward the pineal foramen.
The parietal forms a flat skull table with con- cave lateral margins, owing to a posterior con- striction of the skull table. Distinct yet short an- terolateral processes of the parietal are embraced by the posterior ends of the frontal and of the postfrontal. The relatively large pineal foramen is located close to the anterior margin of the parietal (Fig. 12A). The medial suture, separating the orig- inally paired parietals, is still visible at the ante- rior margin of the pineal foramen. The laterally descending flange of the parietal is distinct. It contributes to the formation of a secondary lateral wall of the braincase as it meets the epipterygoid ventrally and the squamosal posteroventrally. The dermal encrustations on the parietal skull table re- semble those of Placochelys: an unpaired postero- medial one, and two protuberances along each side of the skull table.
The prefrontal is small but relatively broadly exposed in dorsal view by comparison with Pla- cochelys. Located at the anterodorsal margin of the orbit, it contacts the frontal dorsally, the nasal anteriorly, and the maxilla ventrally. It closely ap- proaches but remains excluded from the lacrimal foramen, which is enclosed entirely by the max- illa.
The postfrontal is a broad, plate-like and rough- ly triangular element broadly entering the poster- odorsal margin of the orbit (Fig. 12A). The pos- terolateral margin is deeply concave and angled. The posterior process tapers to a slender tip that lies alongside the parietal, extending backward to the same level as the frontal and separated from the anteromedial margin of the upper temporal fossa by a broad posterodorsal process of the post- orbital.
The postorbital again is a relatively large ele- ment that broadly enters the posteroventral margin of the orbit (Fig. 1 2A). A relatively broad poster- odorsal process meets the parietal at the antero- medial margin of the upper temporal fossa at a level shortly in front of the posterior tip of the postfrontal, but well behind the level of the an- terior margin of the upper temporal fossa. In con- trast to Placochelys, the postorbital only margin-
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
25
Fig. 12. Skull of Cyamodus rostratus Munster (holotype, umo BT 748): A, dorsal view; B, ventral view. Scale bar = 20 mm. For abbreviations, see p. 3.
ally overlaps the laterally descending flange of the parietal and does not extend further backward on the descending process of the parietal than does its dorsally exposed part.
Each maxilla carries two small, bulbous teeth, located at the level of the internal naris (Fig. 12B). Only the left internal naris is preserved, and its posterior margin appears smooth and natural. This indicates that the posterior maxillary tooth is lo- cated in front of the posterior margin of the in- ternal naris. The maxilla enters the anterolateral margin of the internal naris, but the precise lo- cation of its contact with the vomer along the an-
terior margin of the internal naris cannot be as- sessed. The palatine meets the maxilla along the posterolateral margin of the internal naris at a lev- el of about the midpoint of the longitudinal di- ameter of the latter. Behind the internal naris, the maxilla can be followed in ventral view as a ta- pering shelf of bone that meets the jugal well in front of the anterior margin of the subtemporal fossa (Fig. 12B).
In lateral view, the maxilla is seen to broadly enter the posteroventral margin of the external naris (Fig. 13 A). Between the external naris and the orbit, the maxilla forms a distinct yet small
26
FIELDIANA: GEOLOGY
Fig. 12. Continued.
and slender ascending process, wedged in be- tween the nasal and prefrontal. Further back, the maxilla broadly enters the ventral margin of the orbit, enclosing the lacrimal foramen at the antero- ventral corner of the orbit (Figs. 12A, 13 A). Pos- teriorly the maxilla meets the jugal in an essen- tially vertically oriented, sigmoidally curved, and interdigitating suture at a level somewhat behind the posterior margin of the orbit. Unlike in other cyamodontoids, the maxilla contacts the postor- bital at the posteroventral margin of the orbit (Fig. 13 A).
The jugal is incompletely preserved because of breakage of the temporal arch. In lateral view, the
jugal is seen to form a short, small, anterior pro- cess that enters between the postorbital and max- illa but remains excluded from the orbital margin by the contact of the latter two bones (Fig. 13A). In ventral view, the jugal is seen to define the anterior margin of the subtemporal fossa. The right side of the skull shows an anterior palatal process of the jugal that enters deeply between the maxilla and the palatine, reaching anteriorly to a level slightly in front of the anterior palatine tooth plate (Fig. 12B). This corresponds to the level of anterior extension of the postorbital along the pos- teroventral margin of the orbit. Preservation is not good enough to ascertain whether the jugal be-
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
27
fc.st
28
FIELDIANA: GEOLOGY
Table 3. Measurements of the palatine tooth plates of Cyamodus rostratus (holotype, umo BT 748). All measurements in mm; approximate values in parenthe- ses.
|
right palatine |
||
|
longi- tudinal 0 |
trans- verse 0 |
|
|
first toothplate |
8.7 |
8.6 |
|
second toothplate |
9.5 |
(9.8) |
|
third toothplate |
27.5 |
23.2 |
comes exposed in dorsal view in the floor of the orbit between the maxilla and the postorbital lat- erally and the palatine medially, or whether the palatine meets the maxilla and postorbital dorsal to this anterior process of the jugal. In lateral view, the jugal can be observed to form a distinct, tapering posteromedial process that extends along the anteromedial margin of the subtemporal fossa, where it enters between the palatine dorsally and the pterygoid ventrally (Fig. 13 A). A similar pos- teromedial process of the jugal has not been re- corded for other cyamodontoids. The ectoptery- goid would be expected to be located at the an- teromedial margin of the subtemporal fossa, but it is absent in Cyamodus rostratus.
The palatal view of the skull is somewhat dif- ficult to interpret because of damage due to pres- ervation and preparation (Fig. 12B). The vomers are paired elements that separate the internal nares from one another, but other than that little can be said about the precise nature of their contacts to neighboring bones. The palatines are the domi- nant elements in the dermal palate, each carrying three tooth plates, of which the posteriormost plate is much larger than the two anterior ones (Table 3). The nature of the dentition in Cyamo- dus rostratus has generated considerable contro- versy because some of the teeth are obviously glued to the underlying bone surface at places where they may not have been originally located. This is particularly true of the anteriormost left palatine tooth. Kuhn-Schnyder (1965a) recapitu- lated the controversy and concluded that Cyamo- dus rostratus has two premaxillary, two maxil- lary, and three palatine teeth, the same tooth count
proposed by H. v. Meyer (1863). I concur with this conclusion. The palatine bones themselves are badly broken, and the medial suture between them can be only partially identified. The left pos- terior palatine tooth plate is incompletely erupted. The posterior dental lamina foramina are distinct, located posterior to the palatine tooth plates on the palatine-pterygoid suture.
The right pterygoid is incompletely preserved, but the left pterygoid indicates that the length of its palatal exposure is short relative to the length of the palatine. The distance from the posterior margin of the dermal palate to the posterior dental lamina foramen is slightly larger or equal to 10 mm; the distance from the (left) posterior dental lamina foramen to the posterior margin of the left internal naris approximates 58 mm. The ratio of pterygoid length to palatine length thus is approx- imately 0.17.
The (left) pterygoid forms a distinct, longitu- dinally oriented ventral flange that unfortunately is broken along its ventral edge. It is therefore impossible to unequivocally ascertain whether there was a single or a double ventral projection on the pterygoid flange. However, the bipartition of the broken bone surface by a small interme- diate stretch of finished bone with a slightly con- cave surface strongly suggests a double ventral projection, as is also observed in specimen smns 17403 (see below). The left lateral view of the skull shows rather distinctly the anterior extent of the pterygoid along the medial margin of the sub- temporal fossa (Fig. 13 A). The pterygoid extends to a level slightly behind the anterior margin of the posterior palatine tooth plate and approaches the anterior margin of the subtemporal fossa more closely than in other cyamodontoids.
The quadrate has a weakly concave posterior margin and is covered laterally by the quadrato- jugal. The suture separating the quadratojugal from the quadrate and the squamosal is distinct in occipital view; in lateral view, cracks and dermal encrustations obscure the dorsal delineation of the quadratojugal from the squamosal.
The dominant element in the lateral wall of the braincase is the epipterygoid (Fig. 13 A). As in other cyamodontoids, the epipterygoid can be de- scribed as consisting of two parts, a posterior part
Fig. 13. A, Skull of Cyamodus rostratus Miinster (holotype, umo BT 748) in left lateral view. B, Left side of the occiput of Cyamodus rostratus Miinster (holotype, umo BT 748). C, Right side of the occiput of Cyamodus rostratus Miinster (referred specimen, SMNS 17403). Scale bar = 20 mm. For abbreviations, see p. 3.
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
29
with a convex lateral surface and an anterior part with a concave lateral surface, trending toward the midline of the skull. The ventral margin of the posterior part of the epipterygoid is formed by unfinished bone and overhangs the palatoquadrate cartilage recess. The ventral margin of the anterior part is sutured to the dorsal surface of the palatine. The dorsal margin of the epipterygoid meets the laterally descending flange of the parietal. Unlike Placochelys, Cyamodus rostratus shows no dis- tinct ventromedial process of the postorbital abut- ting the lateral surface of the anterior dorsal part of the epipterygoid. Instead, the ventral process of the parietal is exposed as a narrow strip of bone between the posteromedial process of the postor- bital (lining the anteromedial margin of the upper temporal fossa) and the dorsal margin of the epi- pterygoid. Between these two bones, the descend- ing process of the parietal can be followed ante- riorly up to the posterior margin of the foramen interorbitale (paired openings between the orbits filled, in life, by the cartilaginous interorbital sep- tum; Gaffney, 1972), such that the epipterygoid contacts the parietal rather than the postorbital at the posterodorsal corner of the foramen interor- bitale.
The posterior margin of the epipterygoid is deeply concave and forms the anterior margin of the trigeminal incisure. The posterior margin of the trigeminal incisure is formed by the prootic. Behind and dorsal to the trigeminal incisure and above the prootic, the epipterygoid forms a dis- tinct, slender posterior process that meets the squamosal at about the midpoint of the dorsal margin of the posttemporal fossa (Fig. 13 A).
The posterodorsal, posterior, and posteroventral margin of the posttemporal fossa is formed by the squamosal (Fig. 13B). The otic process of the squamosal is shorter in Cyamodus rostratus than in Placochelys; it meets the prootic at about the midpoint of the ventral margin of the posttem- poral fossa, corresponding to the midpoint of the anterior margin of the paroccipital foramen (Fig. 12A; not exposed in lateral view in Cyamodus rostratus). At the same time, the prootic and the otic process of the squamosal form the dorsal margin of the posterior part of the palatoquadrate cartilage recess.
The ventral margin of the palatoquadrate car- tilage recess is formed by the palatine anteriorly and by the quadrate posteriorly. Other than in Pla- cochelys, the palatine does not contact the quad- rate lateral to the recess in Cyamodus rostratus (Fig. 13 A). This results is a greater width of the
palatoquadrate cartilage recess, with the pterygoid forming its lateral margin between palatine and quadrate.
The right side of the occiput is very poorly pre- served in the holotype of Cyamodus rostratus. On the left side (Fig. 13B), the posttemporal fossa is larger than in smns 17403 or in Cyamodus kuhn- schnyderi (see descriptions below). It is possible, however, that the lateroventral margin of the post- temporal fossa is incomplete (broken); parts of the squamosal may be missing, such that the pteroc- cipital foramen forms a deep cleft located in the lateroventral corner of the posttemporal fossa. On the other hand, breakage of the squamosal cannot be unequivocally ascertained, such that the con- figuration of the pteroccipital foramen in Cyamo- dus rostratus may represent the plesiomorphic condition relative to other cyamodontoids (see discussion of the derivation of the cyamodontoid braincase from that of Placodus, below). Other than at the lateroventral corner, the margins of the posttemporal fossa are complete, indicating a rel- atively large size.
The holotype of Cyamodus rostratus shows the distal tip of the posterior dorsal process of the epipterygoid to become exposed at the dorsome- dial corner of the posttemporal fossa in occipital view (as in smns 17403, see below). In addition, the specimen shows the bilaterally symmetrical presence of an additional separate ossification in the occiput, a narrow strip of bone that extends from the medial corner of the posttemporal fossa toward the juncture of supraoccipital and exoccip- ital (Fig. 13B). The ventral suture, separating this element from the opisthotic, is distinct on both sides of the skull; the dorsal suture, separating this element from the supraoccipital, is distinct on the left side of the skull but appears partially fused on the right side. No comparable epiotic ossifi- cations are known in other cyamodontoid skulls.
Kuhn-Schnyder (1965a) described the presence of postparietals and tabulars in the occiput of the holotype of Cyamodus rostratus; however, the supposed suture separating the (dorsal) parietal from the (ventral) postparietal corresponds to the margin of a dermal encrustation situated on the posterior margin of the parietal skull table. The sutures indicated by Kuhn-Schnyder (1965a) to be located between the postparietals and tabulars re- main enigmatic.
The supraoccipital is a relatively broad plate carrying a low median crest and defining the dor- sal margin of the foramen magnum. The lateral margins of the foramen magnum are formed by
30
FIELDIANA: GEOLOGY
the exoccipitals. The basioccipital, forming the occipital condyle, is badly eroded and difficult to separate from the exoccipitals. However, a dis- tinctly thickened rim runs along the ventral mar- gin of the foramen magnum that seems to be bi- partitioned at the midline of the skull. If indeed composed of the exoccipitals, this rim would in- dicate that the latter bones meet dorsal to the ba- sioccipital, as they do in Cyamodus kuhnschny- deri.
The opisthotic forms the paroccipital process, which is severely damaged by a horizontal crack passing through it (Fig. 13B). This crack also ob- scures the vagus foramen. In ventral view, a me- dioventral extension of the squamosal appears su- tured to the anterior aspect of the paroccipital pro- cess (as in smns 17403, see below). The distal end of the paroccipital process expands into a distinct, posteroventrally directed tubercle, as is also ob- served in smns 17403, as well as in Cyamodus kuhnschnyderi (Nosotti & Pinna, 1996). There is, in Cyamodus rostratus, no distinct buttress on the squamosal to receive the distal end of the paroc- cipital process.
Specimen smns 17403 is an incomplete skull that was used by Nosotti and Pinna (1993b) in support of their interpretation of the relations of the squamosal to the quadratojugal in the tempo- ral arch. The specimen is supposed to show a dis- tinctive suture separating the squamosal from the quadratojugal in the posterolateral corner of the upper temporal fossa. This suture starts laterally as a crack, below which the suture cannot be pur- sued on the lateral surface of the temporal arch. On the narrow dorsal surface of the temporal arch, the suture forms a "V" with the apex pointing backward. The lateral shank of that "V" crosses a dermal tubercle, but because these dermal tu- bercles secondarily fuse to the posterolateral as- pect of the temporal arch, one would expect the suture to be concealed by this tubercle rather than passing through it. The medial shank of the suture meets the margin of the temporal fossa, but from there the suture cannot be followed onto the me- dial surface of the temporal arch. It is for these reasons that I consider this supposed suture be- tween squamosal and quadratojugal not to be un- equivocally distinct from a break or an artifact of preparation. By contrast, there is on the lateral surface of the temporal arch, at about the level of the dorsal head of the quadrate and below a pos- terolaterally placed dermal tubercle, a horizontal groove that might likewise be interpreted as a su- ture between the (dorsal) squamosal and the (ven-
Fig. 14. Cyamodus rostratus Miinster (referred spec- imen, smns 17403); double ventral projection of the pterygoid flange. Scale bar = 10 mm.
tral) quadratojugal in a position comparable to that seen in Placochelys.
The right lateral wall of the braincase again shows rather distinctly the posterior dorsal pro- cess of the epipterygoid, which meets the squa- mosal at the anterodorsal margin of the posttem- poral fossa, and from below which emerges the prootic. The ventral view of the skull shows the large posterior palatine tooth plates with the den- tal lamina foramina located posterior and postero- medial to the latter. The longitudinally oriented ventral flanges of the pterygoid are well preserved on both sides of the skull, and both show a dis- tinctly concave ventral margin. This results in a double ventral projection of the pterygoid flange, the anterior one located at the level of the poste- rior dental lamina foramina, the posterior one lo- cated at the posterolateral corners of the dermal palate (Fig. 14).
The most important information that can be ob- tained from specimen smns 17403 relates to the well-preserved right side of the occiput (Fig. 13C). Because of an increased occipital exposure of the parietal, squamosal, and opisthotic, the posttemporal fossa appears reduced by compari- son with the holotype of Cyamodus rostratus and more closely resembles the occiput of Cyamodus kuhnschnyderi, which again shows reduced post- temporal fossae. Because the pteroccipital fora- men is also of conventional size and position (at the lower margin of the posttemporal fossa) in smns 17403, the question arises again whether the lateroventral corner of the posttemporal fossa of the holotype of Cyamodus rostratus was subject to damage. As in the holotype of Cyamodus ros- tratus, however, smns 17403 shows the posterior dorsal process of the epipterygoid gaining an oc-
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
31
Fig. 15. 20 mm.
Cyamodus rostratus Munster (referred specimen, umo BT 2172); lower jaw, occludal view. Scale bar
cipital exposure at the dorsomedial corner of the posttemporal fossa (Fig. 13C).
The right paroccipital process of specimen smns 17403 is intact. Its distal end is expanded into a distinct posteroventral tubercle, a character shared by the holotype of Cyamodus rostratus and by Cyamodus kuhnschnyderi (Nosotti & Pinna, 1996). Again, there is no buttress on the squa- mosal to receive the distal end of the paroccipital process. Instead, the distal end of the paroccipital process (opisthotic) meets the squamosal in a broad, interdigitating suture that trends ventrolat- erally from the lateroventral corner of the post- temporal fossa (Fig. 13C). A similar sutural re- lation of the distal end of the paroccipital process to the squamosal appears to be present in the ho- lotype of Cyamodus rostratus and in Cyamodus kuhnschnyderi (Nosotti & Pinna, 1996).
Specimen umo BT 2172 (Drevermann, 1928, PI. 23, Fig. 2) is a lower jaw with both rami in articulation (Fig. 15) that may be referred to Cy-
Table 4. Measurements of the dentary tooth plates of Cyamodus rostratus (referred specimen, umo BT 2172). All measurements in mm.
|
left |
right |
|||
|
longi- tudinal0 |
trans- verse 0 |
longi- tudinal 0 |
trans- verse 0 |
|
|
anterior dentary tooth |
14.7 |
12.2 |
15.7 |
12.1 |
|
posterior dentary tooth |
34.2 |
27.3 |
35.8 |
26.7 |
amodus rostratus. The bone surface is badly erod- ed, and the tips of the retroarticular processes as well as the apex of both coronoid processes are incomplete. Each mandible carries an anterior tooth of a generally bulbous shape but retaining a transversely oriented anterior crest. This anterior tooth is followed, after a short diastema, by two dentary tooth plates, of which the posterior one is distinctly larger (Table 4). The only unusual fea- ture of the specimen is that the elongated lower jaw symphysis is deeply excavated (hollow) if looked at in posterior view.
Specimen smf R-4040 is a much better pre- served but smaller (Tjuvenile; Drevermann, 1928) left mandible referable to Cyamodus rostratus, which was described by Drevermann (1928, PI. 23, Figs. 3a-d) and Rieppel (1995a, Fig. 31). It shows all the cyamodontoid characteristics, such as the large coronoid process formed by the cor- onoid bone, which closely approaches the ventral margin of the lower jaw, and the posterior dentary tooth plate, which is partially located medial to the coronoid process. The suture between the sur- angular and angular is distinctive in this speci- men, perhaps because of its possible juvenile sta- tus (Rieppel, 1995a, Fig. 31). It shows that in the area behind the coronoid process, the lateral ex- posure of the surangular is larger (deeper) than that of the angular, in contrast to Placochelys (but note the difficulty in delineating the angular-sur- angular suture in the latter specimen). In smf R- 4040, part of the angular-surangular suture is ob-
32
FIELDIANA: GEOLOGY
|
i |
|
|
• |
|
|
V M |
■1 |
|
3 |
Fig. 16. Skull of Cyamodus muensteri (holotype, bsp AS VII 1210): A, dorsal view; B and C, ventral view. Scale bar = 20 mm.
Table 5. Dentitional characters for the species of the genus Cyamodus.
posterior palatine tooth-plate
|
maxilla |
palatine |
long. 0 |
|
|
trans. 0 |
|||
|
C. rostratus |
2 |
3 |
1.19 |
|
C. muensteri |
3 |
2 |
1.32 |
|
C. laticeps |
3 |
2 |
1.38 |
|
C. tarnowinensis |
3 |
2 |
1.41 |
|
C. kuhnschnyderi |
2 |
2 |
1.16-1.29 |
scured by a laterally protruding boss, which may represent a dermal encrustation on the lower jaw (less distinct and perhaps abraded in umo BT 2172). smf R-4040 shows four dentary teeth in- creasing in size from front to back, of which the posteriormost one is again much larger than the three anterior ones. This may represent ontoge- netic tooth variation. Kuhn-Schnyder (1959) de- scribed the ontogenetic reduction of tooth posi- tions in Cyamodus hildegardis, which also affects dentary teeth. The diastema observed between the anterior and the two posterior dentary teeth in umo BT 2172 may thus have formed by reduction of one dentary tooth, the second from the front end of the mandible in smf R-4040.
The Cranial Anatomy of Cyamodus muensteri (Agassiz, 1839)
Agassiz figured and named (Placodus miin- steri) the holotype on plate 71, published as part of volume II of his Recherches sur les Poissons fossiles in 1839; the accompanying text was pub- lished in chapter VI, 2nd part, volume II in 1844 (Brown, 1890). The holotype of Cyamodus muen- steri is a small cyamodontoid skull (bsp AS VII 1210) that is badly preserved and poorly prepared (Fig. 16). The maximum length of the skull is 106.5 mm; the maximum width is 1 15 mm. Large parts of the skull are reconstructed and heavily painted. Agassiz (1833-1845) diagnosed the spec- imen mainly on the basis of its dentition, yet not- ed the unusually short rostrum and the wide tem- poral arches (both reconstructed in plaster, how- ever).
Indeed, the entire rostrum was missing in the original skull. Attempting a reconstruction of the skull, Miinster rounded off its anterior end with
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
33
Fig. 17. Skull of Cyamodus muensteri (Agassiz; holotype of C. "laticeps" Owen, bmnh R 1644): A, dorsal view; B, ventral view. Scale bar = 20 mm. For abbreviations, see p. 3.
plaster (as indicated by a dotted line in the figure published by Munster 1 830) and arranged a series of small teeth (three on either side; the first right "maxillary" tooth is now broken) along its mar- gin, as figured by Agassiz (1833-1845). The re- construction of the skull by Munster is certainly unnatural, as was already noted by Meyer (1863), and the only indication of what the specimen may originally have looked like is Munster's (1830) first illustration. This shows the two enlarged, posterior palatine tooth plates in situ. Dividing the longitudinal diameter of the better preserved right posterior palatine tooth (25.9 mm) by its trans- verse diameter (19.6 mm) yields a ratio of 1.32, which is larger than the equivalent ratio for Cy- amodus rostratus but very close to the values ob- tained for Cyamodus "laticeps" (see below). In front of the left posterior palatine tooth plate is
the smaller, anterior palatine tooth plate, aligned with the left posteriormost maxillary tooth, which is located lateral to it. The right side of the skull shows three marginally positioned teeth increas- ing in size from front to back, as is also the case for the three maxillary teeth of Cyamodus "lati- ceps" (Owen, 1858). As noted by Meyer (1863, PI. 31, Fig. 2), the anteriormost one of these three marginal teeth (now broken) is located slightly more medially than the posterior ones, as is again the case for the anteriormost maxillary tooth of Cyamodus "laticeps." Cyamodus muensteri can therefore be reconstructed to share three maxillary teeth with Cyamodus "laticeps" (and Cyamodus tarnowitzensis Giirich, 1884). Dentitional charac- ters (Table 5) therefore suggest synonymy of Cy- amodus muensteri and Cyamodus "laticeps" (the first name takes priority; the holotype of Cyamo-
34
FIELDIANA: GEOLOGY
B
Fig. 17. Continued.
dus tarnowitzensis Giirich, 1 884, can no longer be located today, and the schematic illustration of the poorly preserved specimen is not diagnostic at the species level).
The Cranial Anatomy of Cyamodus "laticeps" (Owen, 1858)
The holotype of Cyamodus "laticeps" (bmnh R 1644) is an incomplete skull from the Upper Mu- schelkalk (mol) of Bayreuth (Fig. 17). The right posterior lateral part, comprising the skull table and the temporal region, is missing. Owen's (1858, Pis. IX and X) illustrations represent a mir- ror image of the specimen. The skull is incom-
pletely prepared, especially the left side of the braincase and the preserved left side of the occi- put. It also appears that parts of these regions have been reconstructed and smoothed over using a mixture of glue and ground bone substance, as was commonly done with fossils for sale from Bayreuth, but without X-ray analysis, the extent of reconstruction is impossible to discern. The maximal length of the skull, as preserved, is 186 mm; the maximal width is 154 mm.
In size and general appearance, the skull close- ly resembles that of Cyamodus kuhnschnyderi (see below). The basicranial length can be recon- structed to approximate 150 mm in Cyamodus "laticeps," whereas the transverse diameter of the (left) upper temporal fossa may have approximat-
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
35
ed 65 mm. Dividing the basicranial length by the transverse diameter of the upper temporal fossa yields a ratio of approximately 2.3 (i.e., very close to Cyamodus kuhnschnyderi) and distinctly small- er than the ratio in cyamodontoids outside the ge- nus Cyamodus, which all have a relatively nar- rower upper temporal fossa. Cyamodus "lati- ceps" differs from Cyamodus kuhnschnyderi in its dentition, however, as well as in the relation of the nasal and premaxillary. Because of damage done to the bone surface during early preparation, suture lines can be only partially identified on the skull of Cyamodus "laticeps," primarily in the preorbital region of the skull. In the snout, the maxilla appears to meet the premaxilla in a suture that enters the anterolateral corner of the external naris. If correctly identified, this suture forms an anteriorly wide-open V as it trends toward the lat- eral margin of the upper jaw. The maxillary-pre- maxillary suture lies in a distinctly more anterior position in Cyamodus "laticeps" by comparison with Cyamodus rostratus and Cyamodus kuhn- schnyderi (Nosotti & Pinna, 1996). The possibil- ity remains, however, that what appears to be the premaxilla-maxilla suture may, in fact, be a break.
The external nares of Cyamodus "laticeps" are elongate and kidney-shaped. Their lateral (ven- tral) margin is formed almost exclusively by the maxilla. The nasal broadly enters the posterior and posteromedial (posterodorsal) margin of the external naris, whereas the premaxilla forms its anterior and anteromedial (anterodorsal) margin. Other than in Cyamodus kuhnschnyderi, the na- sals are paired in Cyamodus "laticeps," and they are separated from one another by elongate pos- terior (nasal) processes of the premaxillae, which meet the frontal at a level shortly in front of the anterior margin of the orbit (Fig. 17 A). The fron- tal forms distinct anterolateral lappets, which re- main separated from the maxilla by a broad con- tact of the nasal with the prefrontal. In Cyamodus kuhnschnyderi, the nasals are fused and meet the frontal in an almost straight transverse suture at a level shortly behind the level of the posterior mar- gin of the external naris.
The prefrontal of Cyamodus "laticeps'" is dis- tinct on the right side of the skull, located at the anteromedial (anterodorsal) margin of the orbit and excluded from the lacrimal foramen, which is enclosed by the maxilla only. Closely comparable to Cyamodus kuhnschnyderi, a distinct "basior- bital furrow" (Nosotti & Pinna, 1996) lines the lateral (ventral) margin of the orbit, but within it
only two foramina can be identified in Cyamodus "laticeps" (Fig. 17 A). The larger anterior fora- men represents the lacrimal foramen, whereas the smaller posterior one must represent the infraor- bital foramen (sensu Oelrich, 1956), transmitting the infraorbital nerve. Subsequent to preparation and preservation, the elements bordering the bas- iorbital furrow cannot be identified.
Other than these, few sutural details can be gleaned from the dorsal view of the skull. The anterior tip of the left postfrontal can be identified at the posteromedial (posterodorsal) margin of the orbit, which by comparison with the right pre- frontal indicates that these two elements were sep- arated from one another by the frontal along the dorsal margin of the orbit. The frontal is unpaired (fused). The lateral part of the suture separating the maxilla from the jugal is distinct on the left side of the skull, but this suture cannot be traced to the lateral (ventral) margin of the orbit. The posterior tip of the maxilla can be seen to extend to a level somewhat behind the midpoint of the longitudinal diameter of the orbit but in front of the posterior margin of the latter. Extrapolating from the identifiable part of the maxillary-jugal suture, the latter must have lined at least the pos- terior part of the basiorbital furrow. There is a weak indication of the suture separating the post- orbital from the jugal at the posterolateral (pos- teroventral) corner of the orbit, but this could also represent a crack. The left temporal arch is dam- aged, as the dorsoventral compression of the skull resulted in the formation of a deep trough, but the unaltered dorsal (medial) part of the surface pre- serves the interdigitating suture separating the postorbital from the squamosal. Large tubercles have secondarily fused to the surface of the squa- mosal along the posterior and posterolateral mar- gin of the upper temporal fossa.
The lateral braincase wall reveals very little structural detail. The most conspicuous feature is the trigeminal incisure, located between the epi- pterygoid (anteriorly) and the prootic (posterior- ly). The contours of these two elements, as well as the palatoquadrate cartilage recess and the pter- occipital foramen, remain indistinct, which raises the suspicion that at least part of the lateral brain- case wall has been reconstructed. The same is true for the preserved (left) part of the occiput.
The ventral view of the skull reveals even few- er sutural details (Fig. 17B). Identifiable are the ventromedial suture line between palatines, vo- mers, and premaxillae, and the suture between the quadrate ramus of the (left) pterygoid and the
36
FIELDIANA: GEOLOGY
Table 6. Measurements of the tooth plates of Cy- amodus muensteri (holotype of C. "laticeps," bmnh R 1644. All meaurements in mm.
|
kmgi- tudmal0 |
trans- |
|
|
1st pre maxillary |
10.6 |
9 |
|
2nd pre maxillary |
8 |
8.2 |
|
1st maxillary |
11.3 |
10.4 |
|
2nd maxillary |
13.2 |
11.5 |
|
3rd maxillary |
17.2 |
17 |
|
1st palatine |
21.5 |
17.8 |
|
2nd palatine |
44.3 |
33 |
quadrate. The internal nares are relatively small, oval openings that appear to be located a little further posteriorly than in Cyamodus kuhnschny- deri. However, this could also be an impression created by differences of the palatal dentition. The posterior dental lamina foramen is located behind the posterior palatine tooth plate. The suture sep- arating the pterygoid from the palatine can be seen to extend from the lateral corner of the left posterior dental lamina foramen in an anterolat- eral direction to the medial margin of the subtem- poral fossa. There is no indication of the presence of an ectopterygoid. The sutural relations of the palatine and jugal at the anterior margin of the subtemporal fossa remain obscure. Cyamodus "laticeps" shows the well-preserved double ven- tral projections of the longitudinally oriented pter- ygoid flange that are characteristic for the genus (Fig. 17B).
The palatal dentition is well preserved and cor- responds to the description given by Owen (1858). Cyamodus "laticeps" shows two premax- illary teeth (a replacement tooth is located medial to the second tooth on the right premaxilla), whereas the holotype and paratype of Cyamodus kuhnschnyderi show only a single premaxillary tooth. Two premaxillary teeth are preserved in a third, recently described specimen (Rieppel & Hagdorn, 1999), indicating variability of the pre- maxillary dentition in the latter species.
However, all three specimens known of Cy- amodus kuhnschnyderi show two maxillary teeth only, whereas Cyamodus "laticeps" shares with Cyamodus muensteri three maxillary teeth (Fig. 17B). These increase in size from front to back (Table 6), and, as described by Owen (1858), the anteriormost maxillary tooth is located a little
more medially than the succeeding two teeth. Fi- nally, Cyamodus "laticeps" has two palatine teeth, a feature it shares with Cyamodus muensteri and Cyamodus kuhnschnyderi. In all three species, the (smaller) anterior palatine tooth plates, togeth- er with the posterior maxillary teeth, are aligned in a transverse, anteriorly slightly concave row. Dividing the longitudinal diameter of the enlarged posterior palatine tooth plate by its transverse di- ameter results in a ratio of 1 .34, which is distinct- ly higher than the ratio in Cyamodus rostratus and somewhat higher than the ratio in Cyamodus kuhnschnyderi, but very close to the value ob- tained for Cyamodus muensteri.
The Cranial Anatomy of Cyamodus kuhnschnyderi Nosotti and Pinna, 1993a
Cyamodus kuhnschnyderi (smns 15855, 16270) has been the subject of a recent monographic de- scription (Nosotti & Pinna, 1996). The species shares with Cyamodus rostratus the relatively wide upper temporal fossae. Dividing the basicra- nial length (tip of snout to occipital condyle) by the transverse diameter of the upper temporal fos- sa yields a ratio of 2.3 in Cyamodus kuhnschny- deri (smns 15855, with parts of the left temporal arch preserved) and a ratio of approximately 2.45 for Cyamodus rostratus. All other cyamodontoids have relatively narrower upper temporal fenestrae, with a corresponding ratio ranging from 3.4 to 4.95.
The posterior (nasal) processes of the premax- illae meet the nasal at the level of about the mid- point of the longitudinal diameter of the external naris. Unlike all other cyamodontoids included in this study, the nasals are fused in Cyamodus kuhn- schnyderi. In smns 16270, a partially fused suture is still apparent at the posterior margin of the na- sal, but the nasals are clearly fused in their ante- rior part. The nasal broadly contacts the prefron- tal, thus separating the frontal from the maxilla. The frontals are paired but may be partly fused at their posterior margin. They meet the (fused) pa- rietal in a deeply interdigitating suture but remain narrowly excluded from the pineal foramen, which lies in an anterior position within the skull table. As in Cyamodus rostratus, the parietal forms short anterolateral processes, which are em- braced by the frontal and the postfrontal. In ad- dition to these anterolateral processes, each pari- etal forms a more medially located anterior pro- cess, which is embraced by the frontal only. Sim-
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
37
ilar processes are absent in Cyamodus rostratus. The posterolateral margin of the large postfrontal of Cyamodus kuhnschnyderi is deeply concave and angulated. The postfrontal closely approaches the anteromedial margin of the upper temporal fossa, separated from the latter by a narrow con- tact of the parietal with the postorbital.
The lateral view of the skull is very incomplete in both specimens, and the identification of su- tures is rendered difficult by the extensive resto- ration of the specimens. The short jugal indicated in the reconstruction of the skull by Nosotti and Pinna (1996) is almost certainly incorrect and re- flects the incompleteness of the temporal arches in both specimens. The postorbital of Cyamodus kuhnschnyderi does not form a posteromedial pro- cess that overlaps the laterally descending flange of the parietal, as seen in Cyamodus rostratus. The epipterygoid contacts the parietal at the pos- terodorsal margin of the foramen interorbitale. Neither of the specimens is well enough preserved to allow the assessment of whether or not the epipterygoid forms a posterior dorsal process meeting the squamosal at the dorsal margin of the posttemporal fossa. As in Cyamodus rostratus, the palatine fails to meet the quadrate lateral to the palatoquadrate cartilage recess (smns 15855). Other aspects of the lateral wall of the braincase of Cyamodus kuhnschnyderi have been discussed in comparison with Placochelys in the descriptive section above, such as the incomplete ossification of the epipterygoid in both specimens and the possible course of the facial nerve through a fo- ramen at the ventral margin of the prootic, close behind the trigeminal incisure (smns 16270).
In ventral view, the premaxillaries can be seen to remain excluded from the internal naris by a contact of the maxilla and vomer. There is no ev- idence for the presence of an ectopterygoid in Cy- amodus kuhnschnyderi. The palatal exposure of the pterygoid is short relative to the length of the palatine. Dividing the length of the pterygoid (from its posterior margin to the dental lamina foramen on the palatine-pterygoid suture) by the distance from the pterygoid-palatine suture to the posterior margin of the internal naris yields values ranging from 0.2 to 0.3 for Cyamodus kuhnschny- deri. The longitudinally oriented ventral flange of the pterygoid is prominent and bears two distinct ventral projections: the anterior at the level of the posterior dental lamina foramen, the posterior at the posterolateral corners of the dermal palate (smns 15855, right side of skull).
Cyamodus kuhnschnyderi also differs from all
other cyamodontoids with the exception of Mac- roplacus raeticus (Schubert-Klempnauer, 1975) and the specimen smns 17403 (referred to Cy- amodus rostratus above) by a significant reduc- tion in the size of the posttemporal fossa. This results from an expansion of the occipital expo- sure of the parietal, squamosal, and opisthotic. The exoccipitals meet dorsal to the occipital con- dyle in Cyamodus kuhnschnyderi. The right me- totic (jugular, vagus) foramen of smns 16270 pre- serves an internal subdivision by a vertical strut of bone that separates a smaller anterior passage (transmitting the glossopharyngeal nerve?) from a larger posterior division (transmitting the vagus complex?). The paroccipital processes are rela- tively well preserved in specimen smns 15855, which shows the anterior aspect of the distal end of the right paroccipital process to be broadly su- tured to the squamosal. The postero ventral aspect of the distal end of the paroccipital process is ex- panded into a distinct tubercle that projects into the space representing the posterior opening of the cranioquadrate passage.
Cyamodus kuhnschnyderi shows a distinct "basiorbital furrow" (Nosotti & Pinna, 1996). This is a groove running along the ventrolateral margin of the orbit on the inside of the anterior process of the jugal and of the maxilla, bordered medially by the palatine. Along the ventrolateral margin of the orbit, three foramina can be iden- tified. The posteriormost one lies within the basi- orbital furrow, between the jugal and the palatine; the intermediate one lies at the anterior end of the basiorbital furrow, between the maxilla and the jugal; and the anterior one lies at the anteroventral corner of the orbit in front of the basiorbital fur- row, entirely within the maxilla. Nosotti and Pin- na (1996) identified the anterior foramen as the lacrimal foramen and the intermediate one as the infraorbital foramen (sensu Oelrich, 1956), trans- mitting the infraorbital nerve. The interpretation of these foramina proposed by Nosotti and Pinna (1996; see above) leaves the posteriormost fora- men unexplained. I concur with Nosotti and Pinna (1996) that the infraorbital nerve would have come to lie in the basiorbital furrow, but instead of passing across the posteriormost foramen lo- cated in that furrow, it most probably passed through it on its way to the superior alveolar ca- nal. Following this interpretation, the anterior two foramina in the anteroventral corner of the orbit would have transmitted the branches of an ante- riorly bifurcating lacrimal duct. Rather than the three foramina in Cyamodus kuhnschnyderi, there
38
FIELDIANA: GEOLOGY
are only two foramina in the equally distinctly differentiated basiorbital furrow of Cyamodus "laticeps." This indicates some variation in the soft anatomy structures relating to the basiorbital furrow in Cyamodus, which renders the unequiv- ocal identification of the function of these foram- ina difficult if not impossible.
The Cranial Anatomy of Cyamodus hildegardis Peyer, 1931a
Although several skulls of Cyamodus hildegar- dis are available, its cranial anatomy remains very poorly known because of the severe dorsoventral compression of the material (Peyer, 1931a, 1935; Pinna, 1992). The dentition, and its ontogenetic variation, was analyzed by Kuhn-Schnyder (1959). Personal inspection of the available ma- terial of Cyamodus hildegardis did not result in the collection of new data. One of the better pre- served skulls (pimuz T4771) shows large dermal tubercles fused to the posterolateral side of the squamosal, similar to those observed in Cyamo- dus kuhnschnyderi. Doubts have been expressed that the three species of the genus Cyamodus could be treated as congeneric (Kuhn-Schnyder, 1960), and Nosotti and Pinna (1996) have treated the genus Cyamodus as paraphyletic. This prob- lem cannot be solved easily, because the cyamo- dontoid material from the Germanic Muschelkalk consists almost exclusively of skulls, whereas Cy- amodus hildegardis is best known from its post- cranial skeleton. If the three carapace fragments from the Muschelkalk have been correctly as- signed to Cyamodus kuhnschnyderi by Nosotti and Pinna (1996, Fig. 14), they would indicate that the dermal armor of the latter taxon is distinct from that of Cyamodus hildegardis. The enlarged posterior tooth plates of Cyamodus hildegardis are more distinctly elongated than those of the Cyamodus from the Germanic Muschelkalk and approach proportions otherwise typical for the Al- pine genus Psephoderma (Table 7; see also com- ments below).
As incomplete as our current knowledge is of the cranial anatomy of Cyamodus hildegardis, the taxon can be coded for 35.2% of the 54 characters used in the cladistic analysis discussed below. The addition of Cyamodus hildegardis to the analysis did not result in a loss of resolution and corrob- orated, on the basis of cranial characters, the monophyly of the genus Cyamodus, including C. hildegardis.
Table 7. Proportions of the posterior palatine tooth plate throughout the Cyamodontoidea.
|
posterior palatine longitudinal 0 tooth-plate transverse 0 |
|
|
Cyamodus rostratus |
1.19 |
|
Cyamodus kuhnschnyderi |
1.16-1.29 |
|
Cyamodus laticeps |
1.34 |
|
Cyamodus muensteri |
1.32 |
|
Cyamodus hildegardis |
1.33 - approx. 1.4 |
|
Protenodontosaurus italicus |
1.24-1.26 |
|
Phcochelys placodonta |
approx. 1.23- 1.3 |
|
Macroplacus raeticus |
1.41 |
|
Placocheylanus malanchinii |
1.57 |
|
Psephoderma alpinum (juvenile) |
L27 |
|
Psephoderma alpinum (adult) |
1.43 - 1.48 |
|
Placochelys alpis sordidae |
1.44 |
|
Placocheylanus stoppanii |
1.73 |
The Cranial Anatomy of Henodus chelyops v. Huene, 1936
Henodus chelyops is known from a total of eight specimens (comprising seven skulls), all of which come from the same deposit and locality, the upper Gipskeuper (Carnian) of Lustnau near Tubingen, southwestern Germany. Not all the skulls are currently accessible for investigation, because some are mounted on permanent exhibit, and preservation is not equal in all available skulls. The present description is based on the skulls of specimens I and II (syntypes, Huene, 1936), and the skulls of specimens IV and VI, which were collected in 1959 (Fischer, 1959) and have never been the subject of a detailed descrip- tion before (Stein, 1993). Specimen I (Fig. 18) is a skull from which the left mandible has been removed and the right mandible has been left in articulation. This is one of the best specimens for study of the preorbital region of the skull in dorsal view. Although the skull roof is poorly preserved, this specimen shows good detail in the dermal palate and is the best skull for study of the quad- rate suspension and the occiput. Specimen II (Fig. 19A) is the best skull for study of the skull roof and the temporal region behind the orbits. It also provides good detail on the structure of the dermal palate and some information on the occiput. Both
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
39
Fig. 18. Skull of Henodus chelyops v. Huene (syntype, gpit uncatalogued, specimen I of Huene, 1936): A, dorsal view; B, ventral view. Scale bar = 20 mm.
lower jaws have been removed from the skull and preserve most information on the structure of the mandible. The skull of specimen IV (Fig. 19B) shows good detail of the structure of the dermal skull in the pre- and postorbital region. The oc- ciput and palate are rather poorly preserved, as are the mandibles. The skull of specimen VI is generally poorly preserved but provides excellent detail on the nature of the contact between frontal, postfrontal, and parietal.
Henodus is a highly autapomorphic cyamodon- toid in every aspect of its anatomy (Fig. 20). Un- like in all other cyamodontoids, the rostrum of Henodus is extremely short yet broad, and the postorbital segment of the dermatocranium forms a broad, flat covering of the braincase. Both the structure of the dermal palate and its dentition dif- fer markedly from what is seen in other cyamo- dontoids. The skull of Henodus initially creates the false impression of being extremely de- pressed, but closer inspection shows that it is dis- tinctly curved relative to the long axis of the low- er jaw. The preorbital region is steeply inclined relative to the skull table, such that the external
nostrils and the orbits face more or less anteriorly relative to the long axis of the lower jaw.
The premaxillae and maxillae form a short yet broad, spatulate rostrum. Immediately behind the premaxillae the skull is strongly constricted, a trait that corresponds to the rostral constriction observed in other sauropterygians. The skull reaches its narrowest dimension just behind the orbits. As the elements of the skull roof and cheek region together form an essentially horizontal der- mal cover of the skull, the ancestral cheek emar- gination results in a weakly expressed concavity of the lateral margins of the skull table. The skull table reaches its greatest lateral extension just above the quadrate suspension. The posterior mar- gin of the skull table is deeply concave, the squa- mosals projecting backward far beyond the level of the occiput, as is also the case in other cyamo- dontoids.
The paired premaxillae are the principal ele- ments forming the rostrum, and have a complex morphology (Fig. 21 A). Anteriorly the premaxil- lae form a transversely oriented, vertically de- scending flange that terminates in a ventral cutting
40
FIELDIANA: GEOLOGY
Fig. 19. Skull of Henodus chelyops v. Huene. A, Syntype, gpit uncatalogued, specimen II of Huene (1936), dorsal view. B, gpit uncatalogued, specimen IV, skull with lower jaw, left lateral view. Scale bar = 20 mm.
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
41
ppf /P°
Fig. 20. Skull of Henodus chelyops v. Huene. A, Left lateral view, partial reconstruction based on specimen I of Huene (1936). B, Dorsal view, specimen I of Huene (1936). C, Ventral view, reconstruction based on specimens I and II of Huene (1936). Scale bar = 20 mm. For abbreviations, see p. 3.
42
FIELDIANA: GEOLOGY
Fig. 20. Continued.
edge. This cutting edge is reinforced on its ante- rior surface by a series of incompletely individu- alized denticles (Fig. 22), a strip of enamel lining the anterior surface of the cutting edge and mim- icking the presence of a row of minute teeth in the more or less regularly spaced indentations and the development of pulp cavities (Stein, 1993, 1995; Reif & Stein, 1999). In dorsal view the pre- maxilla is exposed as a rather narrow strip of bone that defines the entire anterior margin of the ex- ternal naris and meets the maxilla in the antero- lateral (anteroventral) margin of the external naris, as is also the case in other sauropterygians. Pos- teriorly the premaxillae carry ascending nasal pro- cesses that define the medial margins of the ex-
ternal nares and project posteriorly up to the level of the anterior margin of the orbits, entering be- tween the anterior parts of the frontal s (Figs. 21 A, B).
The nasal is a small, triangular element located at the posteromedial margin of the external naris, its apex pointing posteriorly. Together with the posterior (nasal) process of the premaxilla, the na- sal embraces a well-developed anterolateral pro- cess of the frontal. Posterolaterally, the nasal con- tacts the medial (ascending) process of the maxilla (Figs. 21 A, B).
In dorsal view, the maxilla caps the premaxilla laterally as it forms the lateral margin of the broad, spatulate rostrum. Behind the premaxilla,
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
43
pm n
Fig. 21. Skull of Henodus chelyops v. Huene. A, Dorsal view (specimen IV, gpit uncatalogued). B, Dorsal view (specimen I of Huene, 1936). Scale bar = 20 mm. For abbreviations, see p. 3.
Fig. 22. Premaxillary denticles in Henodus chelyops v. Huene (specimen II of Huene, 1936).
at a level between the external naris and the orbit, the maxilla shows a distinct lateral constriction (Figs. 20, 21). The ascending process of the max- illa is medially directed and forms the posterolat- eral (posteroventral) margin of the external naris before entering between the nasal and prefrontal. The nasal and prefrontal generally remain sepa- rated from one another by a contact of the as- cending process of the maxilla with the anterolat- eral process of the frontal, although this contact may be very narrow (specimen II, Fig. 20B). Pos- teriorly the maxilla is separated from the jugal by a suture that in dorsal view appears V-shaped (the apex pointing backward) and is located at about the level of the midpoint of the longitudinal di- ameter of the orbit (Fig. 20A). The maxilla there- fore defines the anterolateral (anteroventral) mar- gin of the orbit, where it forms a pronounced basi- orbital furrow (Figs. 20A, 21 A, B), otherwise re- corded for Cyamodus kuhnschnyderi and C. "laticeps" only among cyamodontoids (Nosotti & Pinna, 1996). In Henodus, two distinct foramina are located in the anteiror half of the basiorbital furrow, of which the anteiror one may correspond to the lacrimal foramen, whereas the posterior one may have served the passage of the maxillary
44
FIELDIANA: GEOLOGY
branch of the trigeminal nerve. In this character, Henodus resembles Cyamodus "laticeps" more closely than Cyamodus kuhnschnyderi. A lacrimal is absent in Henodus.
The prefrontal is a curved and rather slender element that forms most of the anterior margin of the orbit. It contacts the maxilla anteriorly and laterally (ventrally) and the frontal medially (dor- sally). It always remains separate from the post- frontal along the dorsal margin of the orbit, which is formed by the concave lateral margin of the frontal (Figs. 20B, 21 A).
The frontals are paired elongate and rather slen- der elements. They reach their maximal width at the level of the anterior margin of the orbit. Pos- teriorly, each frontal forms a narrow and elongate posterolateral process, each of which is embraced by equally narrow and elongate anterior processes of the parietal. The details of the contact between frontals and parietal are well preserved in speci- mens II (Fig. 20B), IV (Fig. 21 A), and VI (Fig. 2 IB). The parietal is unpaired (fused). It forms an anteromedial process that enters between the pos- terolateral processes of the frontal. More laterally, the parietal forms a more elongate and equally slender anterior process that enters between the posterolateral process of the frontal and the post- frontal and may extend anteriorly up to a level close to the posterior margin of the orbit.
The postfrontal broadly enters the posterome- dial (posterodorsal) margin of the orbit. Shortly behind the orbit the postfrontal is distinctly con- stricted, which results in an angulation of its lat- eral margin (Fig. 21 A). The posterior process of the postfrontal, which extends between the an- terolateral process of the parietal and the postor- bital, appears broad at its posterior end as it meets the parietal in a deeply interdigitating suture in specimen IV By contrast, the postfrontal tapers to a blunt tip posteriorly in specimens I and II.
The postorbital is a large, broad, platelike ele- ment that defines the posterolateral (posteroven- tral) margin of the orbit and partially invades the space of the upper temporal fenestra (Fig. 20B). The element covers most of the postorbital area of the skull in front of the parietal and is bifur- cated posteriorly as it embraces the tapering an- terior end of the squamosal. The posteromedial process of the postorbital extends along the post- frontal and meets the parietal in a deeply inter- digitating suture. This posteromedial process may (specimen II, Fig. 20B) or may not (specimen IV, Fig. 21 A) exceed the postfrontal in length; it cor- responds to that part of the postorbital that in the
plesiomorphic condition lines the anteromedial margin of the upper temporal fossa. The postero- lateral process of the postorbital may (specimen IV) or may not (specimen II) exceed the postero- medial process in length, and in the plesiomorphic condition contributes to the formation of the tem- poral arch as it lines the anterolateral margin of the upper temporal fossa.
The jugal enters the lateral (ventral) margin of the orbit between the maxilla anteriorly and the postorbital posteriorly. It forms the anterior part of the lateral margin of the (horizontally exposed) cheek before it starts to taper off posteriorly. This posterior process of the jugal may (specimen II, Fig. 20B) or may not (specimen IV, Fig. 21 A, left side of skull) reach as far back as the posterior tip of the postorbital. Specimen IV, however, shows that this character is subject to variation between the left and the right side of the skull.
None of the specimens shows a clear demar- cation of the squamosal from the quadratojugal. In front of the dorsal head of the quadrate, three anterior prongs can be identified within the (hor- izontally exposed) cheek region of the skull. The two medial prongs are here interpreted as repre- senting the squamosal, embracing the posterolat- eral process of the postorbital. The squamosal al- ways remains shorter than the third, lateral prong, which forms the posterior part of the lateral mar- gin of the (horizontally exposed) cheek and which is interpreted as the quadratojugal (by analogy to other cyamodontoids, Fig. 20B). The quadratoju- gal reaches anteriorly to a level close to the pos- terior margin of the orbit (specimen IV, Fig. 21 A). Large osteoderms have secondarily fused to the surface of the squamosal and quadratojugal along the posterolateral and lateral margins of the skull table.
The upper temporal fossa is obliterated in Hen- odus (at least in the adult). Closure of the upper temporal fossa is mainly due to lateral spreading of the parietal ossification in a fan-shaped manner, best seen in specimens II (Fig. 20B) and IV (Fig. 21 A). The only specimen that retains a vestigial upper temporal opening with a smooth, natural margin is specimen II (left side of the skull). The hole in the left side of the skull roof of specimen I (Fig. 18) is certainly an artifact of preservation. At this juncture, a brief comment on the skull of specimen III appears to be in order, because it was claimed by Huene (1938) to show well-developed temporal fossae on both sides of the skull. Al- though difficult to access because it is mounted on permanent exhibit, the skull of specimen III
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
45
allows some observations. The skull is not smaller than that of any of the other specimens. The so- called temporal fenestrae have irregular contours, and these differ on the two sides of the skull, sug- gesting that these temporal openings are the result of breakage, as is unquestionably the case with the opening in the left side of the skull roof in specimen I. In fact, the corresponding area of the skull roof appears relatively weak and prone to breakage in all Henodus skulls. The area of the conjectural temporal fossa is strongly depressed on the right side of the skull in specimen II, and it has been severely damaged in specimens IV and VI. Weakness of this area of the skull roof may result from a lateral thinning of the spreading of the parietal ossification responsible for the oblit- eration of the plesiomorphic temporal fenestra.
Most of the skulls have sustained a longitudinal break, obscuring the pineal foramen and making it difficult to assess whether the parietal is paired or fused. Specimen IV indicates, however, that the parietal is fused at least behind the pineal fora- men. Specimen II shows the elongate and narrow pineal foramen to be located at the anterior end of the parietal (Fig. 20B). The frontal remains ex- cluded from the pineal foramen. It remains un- clear, however, whether the parietal is fused or whether it shows a midline suture in front of the pineal foramen.
The structure of the dermal palate of Henodus is again highly autapomorphic (Fig. 20C). The premaxilla is the dominant element in the broad rostrum in ventral view. The cutting edge of the premaxilla overbites the anterior cutting edge of the dentary. In the lateral part of its ventral sur- face, the premaxilla shows a distinct, transversely oriented trough which receives the lateral part of the cutting edge of the dentary. The broad pre- maxilla appears to enter the anterior margin of the internal naris. None of the specimens shows com- plete sutural contact between the premaxilla and the vomer. A short stretch of an obliquely oriented suture is seen on the left side of the palate of specimen II, located between the internal nares and indicating that the vomer enters the medial margin of the latter.
The maxilla reaches its maximum width in its anterior part, where it forms a lateral projection of the rostrum and enters the lateral margin of the internal naris. The ventral surface of the maxilla carries the highly characteristic, sigmoidally curved groove (Figs. 20C, 23), which was be- lieved by Huene (1936) to have carried keratinous structures similar to baleen. Huene (1936) even
Fig. 23. Right side of dermal palate in Henodus chelyops v. Huene (specimen I of Huene, 1936).
claimed to have prepared such structures from the maxilla of one specimen, but this piece can no longer be located today.
None of the specimens shows in ventral view the sutural separation between the jugal and the maxilla, which in dorsal view is located at the level of the midpoint of the longitudinal diameter of the orbit. The suture between the maxilla and the palatine is distinct, however, running from the posterolateral corner of the internal naris posteri- orly and slightly laterally and terminating at about the posterior two-thirds of the maxillary groove (Fig. 20C). In none of the specimens is it possible to assess whether the posterior third of this groove is still formed by the slender maxilla extending backward between the jugal and palatine or whether this groove continues behind the maxilla either on the jugal or on the palatine.
There is no evidence for the presence of an ec- topterygoid in Henodus.
The palatine is an elongate and rather slender element that is broadest in its anterior part, where it forms the posterior margin of the internal naris. A single small palatine tooth plate is located on
46
FIELDIANA: GEOLOGY
Fig. 24. Suspension of left quadrate in Henodus chelyops v. Huene (specimen I of Huene, 1936). Scale bar = 20 mm.
the posterior part of the palatine, at the back end of the maxillary groove (Fig. 20C). The dental lamina foramen is located posterior and postero- medial to the palatine tooth plate.
The paired vomers are broad elements that form the medial margins of the internal nares, from where they extend backward to meet the elongated pterygoids. The vomers separate the an- terior parts of the palatines from one another. Un- like in any other cyamodontoid, the posterior parts of the palatines are separated by the paired ptery- goids (Figs. 20C).
At the level behind the palatine tooth plate and its dental lamina foramen, the pterygoid extends laterally to cover the entire width of the dermal palate. The pterygoid also forms a weakly ex- pressed, longitudinally oriented flange posterior to the palatine tooth plate with a single ventral pro- jection. The short quadrate ramus of the pterygoid emerges from above this ventral flange and ex- tends posterolaterally to contact the quadrate. In contrast to other cyamodontoids, the pterygoids broadly extend anteriorly between the palatines and meet the vomers at approximately the mid- point of the length of the palatines (Fig. 20C).
The quadrate and its suspension are again of a peculiar structure in Henodus, best observed in specimen I (Figs. 20A, 24). The mandibular con- dyle is transversely expanded and bipartite, to match the saddle-shaped surface of the mandibu- lar articulation. The medial articular facet on the mandibular condyle is distinctly larger than the lateral facet. The lateral surface of the shaft of the quadrate is covered by a descending process of dermal bone, presumably part of the quadratoju- gal. Below the skull table, the quadrate expands
into a large, elongated, posterior (suprastapedial) process, accentuating the concavity of its poste- rior margin. The posterior tip of that dorsal ex- pansion of the quadrate abuts a distinct flange de- scending from the ventral surface of the squa- mosal. Between the dorsal expansion of the quad- rate, the skull roof, and the descending flange of the squamosal there persists, in all specimens of Henodus, a gap that leads into the temporal va- cuity of the skull.
None of the specimens of Henodus shows a well-preserved occiput, but some information can be obtained from specimens 1 and II. The post- temporal fossae are large (i.e., not reduced as in Cyamodus kuhnschnyderi or Macroplacus). The exoccipitals meet dorsal to the basioccipital in the occipital condyle. The sutural contact of the ex- occipital to the supraoccipital, to the basioccipital lateral to the occipital condyle, and to the opis- thotic remains indistinct. As in other cyamodon- toids, the opisthotic forms a distinct ventral flange, which together with the basioccipital tuber enclosed the passage of the internal carotid. The ventral opisthotic flange remains separate from the basioccipital tuber as well as from the basi- cranium. There is evidence for the expansion of the distal end of the paroccipital process into a ventrally directed tuber (now broken), comparable to the same character observed in Cyamodus kuhnschnyderi.
The distal end of the paroccipital process abuts the medial surface of the flange descending from the lower surface of the squamosal. On the left side of the skull of specimen II, this ventral pro- cess of the squamosal expands into a ventral flange that meets a dorsal flange originating from the quadrate ramus of the pterygoid. The squa- mosal and pterygoid thus define the closed lateral margin of the posterior opening of the cranio- quadrate passage, but at the same time conceal the point of articulation of the stapes with the quad- rate in the posterior or ventral view of the skull.
The lateral braincase wall is rather poorly pre- served in all specimens available. As in other cy- amodontoids, the most prominent element is the broad epipterygoid. Details of its relation to neighboring elements cannot be ascertained. A pteroccipital foramen can be identified in speci- men I, but the sutures between the surrounding elements again remain indistinct. A palatoquad- rate cartilage recess is present but still filled with matrix on the left side of the skull of specimen I. It is bordered ventrally by the palatine and pter- ygoid and posteriorly by the quadrate. It remains
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
47
ang
sang ang
sang
Fig. 25. Lower jaw of Henodus chelyops v. Huene (partially reconstructed, based mainly on specimens I and II of Huene, 1936). A, Lateral view; B, medial view; C, dorsal view. Scale bar = 20 mm. For abbreviations, see p. 3.
unclear whether the palatine contacts the quadrate lateral to the palatoquadrate recess. The ventral margin of the palatoquadrate cartilage recess is complete, however, and does not show the gap that exists in these cyamodontoids where the pal- atine remains separate from the quadrate.
Specimens II and III show well-developed and well-ossified hyoid elements overlying the poste- rior part of the dermal palate.
The lower jaw of Henodus is unusually deep and massive (Fig. 25). The anterior end of the dentary is sharply turned inward to form a trans- versely oriented cutting edge. No toothlike struc- tures have ever been identified on the dentary. The mandibular symphysis is narrow and delicate, presumably a secondary development correlated with the development of a spatulate snout.
Behind the anterior, transversely oriented cut- ting edge the dentary carries a sigmoidally curved groove, the counterpart of the maxillary groove (Fig. 25C). At the posterior end of the dentary, behind that groove, is located the single and rel- atively small dentary tooth plate, opposing the palatine tooth plate with an occlusal surface that tilts somewhat medially.
The coronoid process is less well developed in
Henodus compared with other cyamodontoids, which reflects a lesser degree of durophagy in this genus. As in other cyamodontoids, the coronoid process is formed by the coronoid only, which also defines the anterior margin of the adductor fossa. Ventrally, the coronoid expands across the lateral surface of the mandible in a fan-shaped manner, but not to the same degree as is observed in other cyamodontoids, so that it remains rather broadly separated from the ventral edge of the lower jaw (Fig. 25 A). The right mandible of spec- imen II shows a relatively broad contact between the dentary and the coronoid dorsally and be- tween the splenial and angular ventrally. Both these latter elements gain a broad exposure on the lateral surface of the lower jaw. The lateral sur- face of the mandible furthermore shows a char- acteristic relief in that the surangular and angular form a distinct vertical step at the level of the anterior end of the adductor fossa. Superficial jaw adductor muscle fibers must have glided across the dorsal margin of that step (formed by the sur- angular) as they expanded to insert into the lateral surface of the coronoid, surangular, and dentary. The ventral margin of the mandible, composed of
48
FIELDIANA: GEOLOGY
splenial and angular, is distinctly sculpted and un- even in all specimens of Henodus.
The adductor fossa is a deep yet anteroposte- riorly relatively short trough located medial to and behind the coronoid process. Anteriorly, the ad- ductor fossa extends to a level well in front of the posterior margin of the dentary tooth plate.
The articular facet of the mandibular joint clos- es the adductor fossa posteriorly. The articular surface itself is strongly saddle-shaped and bicon- cave to accommodate the mandibular condyle of the quadrate. The chorda tympani foramen is dis- tinct and located on the dorsal surface of the re- troarticular process just behind the articular facet (Fig. 25C).
The retroarticular process is very prominent. It is deep, its dorsal surface slants posteroventrally, and it is distinctly sculpted to facilitate the attach- ment of muscle fibers or tendons.
The medial surface of the lower jaw (Fig. 25B) is covered by the prearticular, angular, splenial, and dentary. The prearticular and angular meet the dentary and splenial in an almost vertically ori- ented suture at the level of the coronoid process. The dentary and splenial close Meckel's canal in medial view. The canal opens anteriorly just be- hind the medial bent of the anterior end of the dentary.
The Cranial Anatomy of Macroplacus raeticus Schubert-Klempnauer, 1975
Macroplacus raeticus is based on an incom- plete skull (bsp 1967 I 324; Fig. 26) from the Ba- varian Alps (Rhaetian Koessen-Formation, Hin- terstein bei Hindelang im Allgau: Schubert- Klempnauer, 1975). The skull is relatively large (maximal length as preserved: 181 mm; maximal width: 180.6), with a broad and relatively high temporal region. The rostrum is missing, with an oblique anterior break passing through the left ex- ternal naris and just in front of the right external naris. The skull was subject to considerable ero- sion. It was prepared with acid and subsequently heavily coated with varnish. This, together with a tendency toward fusion of the bones, renders the unequivocal identification of sutures difficult if not impossible in some areas of the skull. The most prominent feature of the skull is the enor- mous posterior palatine tooth plates. Their dis- tinctly elongate shape may have been the reason why Pinna (1978) considered Macroplacus a ju-
nior synonym of Psephoderma {Psephoderma raeticus).
Unfortunately, the most diagnostic feature of the skull, the rostrum, is incomplete in Macropla- cus. In Psephoderma, the rostrum carries paired grooves on its lower surface that lead up to the internal nares, and the maxilla carries a distinct anterior process entering the rostrum in ventral view along the lateral margin of this ventral groove. In Macroplacus, the ventral surface of the preserved proximal part of the rostrum is deeply concave, forming a single longitudinal groove or trough (Fig. 26C). Likewise, the maxilla does not extend as far anteriorly along the lateroventral margin of the rostrum as it does in Psephoderma. Finally, the proportions and shape of the upper temporal fossae of Macroplacus are distinctly dif- ferent from those of Psephoderma alpinum, which shows elongated but relatively narrow temporal fossae (see Table 8).
The dorsal view of the skull of Macroplacus (Fig. 26B) shows greatly enlarged posterior (na- sal) processes of the premaxillae, that define the entire dorsal (medial) margin of the external nares and extend backward to the level of the anterior margin of the orbits, where they meet the frontals in an interdigitating suture. The nasals are rela- tively small, triangular elements that define the posterior margin of the external nares, and remain separated from one another by the broad posterior (nasal) processes of the premaxillae. The nasals narrow toward their posterior ends as they meet the prefrontals, thus separating the frontal from the maxilla. The posterior tip of the nasal lies at the same level as the posterior tip of the nasal process of the premaxilla. The two bones embrace a relatively small yet distinct anterolateral process of the frontal.
The maxilla forms most of the ventral and pos- teroventral margin of the external naris (Fig. 26A). Unlike in Placochelys and Cyamodus, the maxilla of Macroplacus does not expand medially to floor the external naris. Between the external naris and the orbit, the maxilla forms a distinct and pointed ascending process that is embraced by the nasal (anteriorly) and the prefrontal (pos- teriorly). Further back, the maxilla enters the an- teroventral margin of the orbit. The lacrimal fo- ramen is located in the anteroventral corner of the orbit and is fully enclosed by the maxilla (Fig. 26B). Posteriorly, the maxilla meets the jugal at the level of the midpoint of the longitudinal di- ameter of the orbit. Immediately below the ventral margin of the orbit, the suture separating the max-
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
49
Fig. 26. Skull of Macroplacus raeticus Schubert-Klempnauer (holotype, bsp 1967 I 324): A, left lateral view, B, dorsal view; C, ventral view. Scale bar = 20 mm. For abbreviations, see p. 3.
50
FIELDIANA: GEOLOGY
Fig. 26. Continued.
Table 8. Skull proportions of cyamodontoid placodonts. Measurements based on the holotype except ***), which is msnm V471. Abbreviations: basicranial, distance from tip of snout to occipital condyle; long., longitudinal; temp.f., upper temporal fossa; trans., transverse; *) approximate values as reconstruction of the skull or severe preservational distortion of the skull is involved; **), measurements based on the right side of the skull.
|
basicranial long. 0 temp.f. |
basicranial |
long. 0 temp.f. long. 0 orbit |
|
|
trans. 0 temp.f. |
|||
|
Cyamodus rostratus* |
1.76 |
2.46 |
2.22 |
|
Cyamodus kuhnschnyderi* |
1.55 |
2J |
2.5 |
|
Cyamodus ktticeps* |
1.97 |
2.3 |
2.46 |
|
Cyamodus hildegardis* |
2.2 |
4 |
2 |
|
Macroplacus |
1.72 |
||
|
Placochetys placodonta |
2.28 |
3.62 |
1.94 |
|
Protenodontosaurus italicus** |
2.07 |
3.38 |
1.65 |
|
Psephoderma alpinum*** |
2.46 |
4.95 |
1J9 |
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
51
ilia from the jugal is V-shaped, the apex pointing backward. More ventrally, the maxillary-jugal su- ture becomes vertically oriented and interdigitat- ing. It curves around the ventral margin of the skull shortly behind the posterior maxillary tooth (Fig. 26A).
As in all other cyamodontoids, the prefrontal is located at the anterodorsal margin of the orbit and displays limited dorsal exposure. It descends far down along the anterior margin of the orbit, close- ly approaching but not quite reaching the lacrimal foramen.
The delineation of the postfrontal and postor- bital is difficult. The figures of the specimen (Figs. 26 A, B) therefore entail an element of re- construction. On the right side of the skull, a dis- tinct, slightly interdigitating suture can be fol- lowed from the posterodorsal (posteromedial) margin of the orbit in a posterolateral direction at first, before the suture abruptly turns in a postero- medial direction, continuing its course up to the level of the anterior margin of the upper temporal fossa, where the suture turns laterally and be- comes more deeply interdigitating. By compari- son with other cyamodontoids, this suture must demarcate the medial and posterior margin of the right postfrontal, the latter establishing a medial contact with the frontal and a posterior contact with the parietal.
A distinct and interdigitating suture running into the posteroventral corner of the left orbit must delineate the anteroventral tip of the left postorbital. From the orbit this suture trends in a posterodorsal direction as it extends into the tem- poral arch. As it approaches the upper margin of the temporal arch, it forms a distinct vertical step, but then continues horizontally below the dorsal margin of the temporal arch before entering the margin of the temporal fossa at a level well be- hind the midpoint of the longitudinal diameter of the upper temporal fossa. This suture delineates the posterior lateral process of the postorbital, which in Macroplacus reaches far back within the temporal arch, as it also does in Placochelys and Psephoderma.
What remains unclear on both sides of the skull is the suture separating the postorbital from the postfrontal within the postorbital arch. Along the posterior margin of the left orbit there is a rather distinct step that might indicate the anteroventral tip of the postfrontal. Beyond that, however, it re- mains unclear whether the posterolateral margin of the postfrontal was smoothly curved or deeply convex and angulated, and whether the postfrontal
was broadly or narrowly separated from the an- teromedial margin of the upper temporal fossa by a contact of the postorbital with the parietal (the suture in the drawing represents a reconstruction without observational basis).
The jugal meets the maxilla at the level of the midpoint of the longitudinal diameter of the orbit. It therefore defines the posterior half of the ventral margin of the orbit. In Macroplacus, the jugal does not extend anteriorly beyond the level of the midpoint of the longitudinal diameter of the orbit as it does in Placochelys, where the jugal closely approaches the lacrimal foramen. Whether the ju- gal forms a distinct ventral expansion behind the maxilla as in other cyamodontoids cannot be es- tablished because of breakage. The jugal extends into the temporal arch to a level in front of the midpoint of the longitudinal diameter of the orbit, distinctly less far posteriorly than the postorbital. The posterior end of the jugal tapers to a blunt tip; this tip and the postorbital together embrace the anterior end of the squamosal (Fig. 26 A). Along the ventral margin of the temporal arch, the jugal meets the quadratojugal at a level well be- hind the anterior margin of the upper temporal fossa, that is, further back than in Placochelys.
Macroplacus is an important specimen for re- constructing the relations of the quadratojugal and squamosal within the temporal arch. The V- shaped suture (apex pointing anteromedially) that separates the squamosal from the parietal along the posteromedial margin of the upper temporal fossa is distinct on the left side of the skull. No suture can be observed on either side of the skull that would separate the squamosal from the quad- ratojugal at the posterolateral margin of the tem- poral fossa, as would be required on the basis of Nosotti and Pinna's (1993b) reconstruction of the temporal region of cyamodontoids (see also Pin- na, 1989). Instead, the left side of the skull shows a horizontal line or groove that extends backward from the posterior end of the jugal and appears to represent the suture separating the dorsal squa- mosal from the ventral quadratojugal within the temporal arch (Fig. 26A). More posteriorly, this suture disappears below dermal encrustations on the posterolateral aspect of the temporal arch.
The frontals are paired, although the suture be- tween the two elements may be partially ob- scured. Between prefrontal and postfrontal, the frontal forms the concave dorsal margin of the orbit. Unlike in any other cyamodontoids, the frontal widens conspicuously behind the orbit be- cause of the concave medial margin of the post-
52
FIELDIANA: GEOLOGY
frontal (Fig. 26B). The frontoparietal suture re- mains obscure, which makes it impossible to es- tablish whether or not the frontal enters the pineal foramen.
The large pineal foramen is located at the level of the anterior margin of the temporal fossa. A longitudinal crack runs through the parietal skull table close to its midline, but fusion of the pari- etals is indicated by the absence of any trace of a suture either at the anterior or at the posterior mar- gin of the pineal foramen. The parietal skull table is slightly constricted at its posterior end. Dermal encrustations are present, but ill-defined and weakly expressed on the skull table.
The ventral view of the skull (Fig. 26C) dis- plays the dentition of Macroplacus, which com- prises two maxillary teeth and two palatine tooth plates. The right maxilla is well delineated in ven- tral view. Although shorter than in Placochelys, Protenodontosaurus, and, especially Psephoder- ma, the maxilla carries an anterolateral process that tapers off along the lateral margin of the ros- trum— unlike in Cyamodus, where the maxillary- premaxillary suture is transversely oriented (C. kuhnschnyderi) or even trends in a posterolateral direction (C rostratus). Unfortunately, the ante- rior delineation of the vomers is difficult, which makes it impossible to unequivocally assess whether the premaxilla enters the internal naris in Macroplacus. The suture separating the maxilla from the palatine enters the lateral margin of the internal naris at about the level of the midpoint of its longitudinal diameter. From there the suture curves around the anterior palatine tooth plate and then trends posterolaterally toward the pointed posterior tip of the maxilla, which remains ex- cluded from the anterior margin of the subtem- poral fossa by a contact of the palatine with the jugal. The maxilla carries two tooth plates, the posterior one somewhat larger than the anterior one (Table 9). A nutritive foramen is located me- dial to the anterior maxillary tooth entirely within the maxilla. The nutritive (dental lamina) foramen of the posterior maxillary tooth plate is located behind the tooth on the maxilla-palatine suture.
At the level of the anterior margin of the sub- temporal fossa, the palatine forms a lateral pro- cess that meets a medial process of the jugal (Fig. 26C). The jugal reaches to the anteromedial cor- ner of the subtemporal fossa but does not extend backward along the medial margin of the latter as it does in Cyamodus rostratus.
The palatines are paired, much enlarged ele- ments, each carrying the hypertrophied posterior
Table 9. Measurements of the maxillary and pala- tine tooth plates of Macroplacus raeticus (holotype, bsp 1967 I 324). All measurements in mm.
|
left |
right |
|||
|
tudinal0 |
trans- verse 0 |
longi- tudinal 0 |
trans- verse 0 |
|
|
anterior maxillary tooth |
- |
■ |
15.0 |
15.6 |
|
posterior maxillary tooth |
- |
- |
22.6 |
16.9 |
|
anterior palatine tooth |
217 |
19.8 |
21.0 |
20.7 |
|
posterior palatine tooth |
68.5 |
483 |
- |
palatine tooth plate along with a smaller, anterior palatine tooth (Table 9, Fig. 26C). The available specimens of Psephoderma indicate a relative size increase in the posterior palatine tooth plates dur- ing ontogeny. Given the relatively large size of the skull of Macroplacus, the hypertrophy of the posterior palatine tooth plates may be the result of their positive allometric growth. The right pos- terior palatine tooth plate is missing. A posterior dental lamina foramen cannot be identified behind or within its wide tooth socket. The left posterior dental lamina foramen is represented by a narrow groove, located on the palatine-pterygoid suture posteromedial to the posterior palatine tooth plate.
Because of the large size of the palatine, the pterygoid gains a narrow palatal exposure only. Dividing the distance from the posterior margin of the (left) pterygoid to the left dental lamina foramen (13 mm) by the distance from the left dental lamina foramen to the posterior margin of the internal nares (76.5 mm) yields a quotient of 1.7. The pterygoid forms a distinct longitudinally oriented ventral flange, but erosion on both sides of the skull renders it impossible to assess wheth- er they formed a single or double ventral projec- tions. Along the medial margin of the subtemporal fossa, the pterygoid extends anteriorly to the level of the posterior third of the longitudinal diameter of the posterior palatine tooth plate. There is no evidence for the presence of an ectopterygoid in Macroplacus. The quadrate ramus of the ptery- goid is short and stout, and meets the quadrate in an interdigitating suture. The articular surface on the mandibular condyle of the quadrate is bicon- cave, matching the saddle-shaped surface of the mandibular articulation.
The lateral wall of the braincase is poorly pre- served on both sides of the skull. A number of interesting observations can nevertheless be re- corded. The palatoquadrate recess is distinct, but
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
53
Fig. 27. Skull of Protenodontosaurus italicus Pinna (holotype, mfsn 1819GP): A, right lateral view; B, dorsal view; C, ventral view. Scale bar = 20 mm. For abbreviations, see p. 3.
54
FIELDIANA: GEOLOGY
Fig. 27. Continued.
as in Cyamodus, the palatine fails to contact the quadrate at the lateral margin of the latter (Fig. 26B). This allows the pterygoid to enter the lateral margin of the palatoquadrate cartilage recess be- tween palatine and quadrate. The exact contours and relations of the epiterygoid and prootic are difficult to establish, although the trigeminal in- cisure is distinct on the left side of the skull. The right side of the skull shows a distinct medio- ventral process of the postorbital that, as in Pla- cochelys but unlike Cyamodus, abuts the lateral aspect of the epipterygoid at the posterodorsal corner of the foramen interorbitale.
The posttemporal fossa is even more reduced in Macroplacus than in Cyamodus kuhnschnyderi by the expansion of the occipital exposure of the parietal, squamosal, and opisthotic. The relations
of the epipterygoid and squamosal along the dor- sal margin of the posttemporal fossa remain ob- scure. The ventral margin of the posttemporal fos- sa, which coincides with the anterior margin of the pteroccipital foramen, is formed by the squa- mosal in the posterior half and by the prootic in the anterior half. The otic process of the squa- mosal therefore remains relatively short in Macro- placus, similar to that of Cyamodus rostratus, but unlike that of Placochelys, where it forms the en- tire ventral margin of the posttemporal fossa and extends beyond the anteromedial corner of the lat- ter.
The occiput of the skull of Macroplacus is bad- ly eroded. Only rudiments of the paroccipital pro- cesses can be identified, with their distal tips su- tured to the squamosal. Basioccipital, exoccipi-
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
55
tals, and supraoccipital are all missing. A gap be- tween the rudiments of the left opisthotic and squamosal represents the pteroccipital foramen. As preserved, this foramen is located behind and below the squamosal-prootic bridge, which de- fines the ventral margin of the posttemporal fossa in lateral view. With respect to this character, Ma- croplacus resembles Cyamodus rostratus, but as in the specimen of the latter taxon it remains un- clear to what degree this character is the result of erosion or breakage.
Macroplacus is unique among cyamodontoids in that a foramen pierces the shaft of the quadrate just above the mandibular condyle (exposed in posterior view: Schubert-Klempnauer, 1975, PI. 5, Fig. 3). This foramen has smooth edges and is bilaterally symmetrical. Its function remains un- known, as does the function of the foramen that pierces the suspensorium between quadrate, quad- ratojugal, and squamosal in Placochelys.
The Cranial Anatomy of Protenodontosaurus italicus Pinna, 1990b
Protenodontosaurus italicus is known from two skulls (mfsn 1819GP, 1923GP) from the Carnian of Chiout Zuguin east of Dogna, Udine, north- eastern Italy (Pinna, 1990b). These skulls were recently the subject of a detailed description by Nosotti and Pinna (1999). The skull of Proteno- dontosaurus (Fig. 27) is distinctly higher than that of other cyamodontoids. The taxon differs from all other cyamodontoids by the presence of a sin- gle maxillary tooth (Fig. 27C). The persisting maxillary tooth is located lateral to the anterior palatine tooth, which is the position of the pos- teriormost maxillary tooth in all other cyamodon- toids except Cyamodus rostratus. This indicates that Protenodontosaurus has reduced the anterior maxillary tooth (teeth) and retained the posteri- ormost one for biomechanical reasons. Reduction of the maxillary dentition leaves a distinct diaste- ma separating the maxillary from the premaxil- lary dentition (Nosotti & Pinna, 1996). A similar diastema, but less distinctly developed, is present in those specimens of Cyamodus kuhnschnyderi that retain a single pair of premaxillary teeth only (character 8 of Nosotti & Pinna, 1996: 35). In a third specimen of the same species (mhi 1294; Rieppel & Hagdorn, 1999) with two pairs of pre- maxillary teeth, no diastema is present between premaxillary and maxillary teeth. Given the lack of premaxillary teeth in Placochelys and Pse-
phoderma, a diastema separating premaxillary and maxillary teeth is autapomorphic for Proten- odontosaurus among cyamodontoids and is cor- related with the reduction of the maxillary denti- tion.
The premaxillaries form a short and rounded rostrum. The posterior nasal processes of the pre- maxillaries are rather short and meet the nasal be- tween the external nares in a V-shaped suture (with the apex pointing backward). The nasals are paired, slender, triangular elements that form the entire dorsal (medial) margin of the external na- res. The tapering posterior tips of the nasals ex- tend backward to a level behind the anterior mar- gin of the orbits. A narrow contact of prefrontal and nasal separates the frontal from the maxilla. A distinct anterolateral process of the frontal is embraced by the prefrontal and the nasal (Fig. 27B).
The prefrontal is located rather high on the an- terodorsal margin of the orbit and has a relatively narrow dorsal exposure. It descends along the an- terior margin of the orbit but does not approach the anterior tip of the jugal as closely as in Pla- cochelys.
The maxilla meets the premaxilla at the anter- oventral (anterolateral) margin of the external nar- is. It forms all of the ventral (lateral) and posterior margin of the external naris. Because of the skull proportions of Protenodontosaurus, the maxilla is almost as high as it is long (Fig. 27 A). The dis- tinct ascending process narrowly enters the anteroventral (anterolateral) corner of the orbit, and its pointed tip is embraced by the prefrontal and maxilla.
In the anteroventral margin of the orbit, two foramina can be identified, both fully enclosed by the maxilla (Fig. 27B). The larger foramen is lo- cated medial to the ventral marginal rim of the orbit and represents the lacrimal foramen. The smaller foramen lies on the lateral aspect of the maxilla, that is, lateral to the ventral marginal rim of the orbit. The infraorbital foramen (sensu Oel- rich, 1956), transmitting the infraorbital nerve would have to lie inside the orbit, but it cannot be located in Protenodontosaurus (Nosotti & Pin- na, 1996). Nosotti and Pinna (1998) concluded that the infraorbital nerve must have passed me- dial to the preorbital bridge formed by the pre- frontal, maxilla, and palatine. This interpretation leaves the lateral foramen located within the max- illa unexplained, which raises the question of whether Protenodontosaurus had an anteriorly bi- furcating lacrimal duct, the branches of which
56
FIELDIANA: GEOLOGY
passed through paired foramina in the maxilla at the an tero ventral corner of the orbit. Cyamodus kuhnschnyderi is unique in that it shows three fo- ramina along the lateroventral margin of the orbit, and a bifurcating lacrimal duct may be hypothe- sized to have been present in this taxon (see above). Although one of the two lacrimal foram- ina is not in exactly the same topological position with respect to surrounding bones and the orbit in the two taxa, a bifurcating lacrimal duct might have been a character shared by Protenodonto- saurus and Cyamodus kuhnschnyderi. Alterna- tively, Protenodontosaurus might have had a sim- ple lacrimal duct passing through the larger fo- ramen located on the inside of the orbital rim, whereas the smaller foramen located on the max- illa, close to but lateral to the orbital rim, might represent an unusually high placement of the pos- teriormost superior alveolar foramen.
The vertical portion of the interdigitating suture between maxilla and jugal lies at about the level of the midpoint of the longitudinal diameter of the orbit in Protenodontosaurus (Fig. 27 A), compa- rable to what is seen in Macroplacus but in a somewhat more anterior position than is observed in Placochelys, in which the vertical suture be- tween maxilla and jugal lies behind the level of the midpoint of the longitudinal diameter of the orbit.
The frontals of Protenodontosaurus are paired and meet the parietal in a deeply interdigitating suture at a level between the orbit and the upper temporal fossa (Fig. 27B). The orbital margin of the frontal, between pre- and postfrontal, is straight rather than concave. The frontal closely approaches but does not enter the pineal foramen (Nosotti & Pinna, 1998). The postfrontal is a broad, triangular element with a weakly concave and evenly curved posterolateral margin. It re- mains broadly separated from the upper temporal fossa by the postorbital, which meets the parietal at the anteromedial margin of the upper temporal fossa (Fig. 27B). A broad separation of the post- frontal from the upper temporal fossa is also ob- served in Psephoderma and, to a lesser extent, in Cyamodus rostratus, but not in Placochelys or Cyamodus kuhnschnyderi, where the contact of postorbital and parietal is narrow. The postorbital broadly enters the postero ventral margin of the orbit. It extends backward into the temporal arch, forming the anterior and anterolateral margin of the upper temporal fossa. The posterior tip of the postorbital remains restricted to a level in front of
the longitudinal diameter of the upper temporal fossa.
The jugal broadly enters the ventral margin of the orbit and closely approaches but does not en- ter the lacrimal foramen (Fig. 27A). Posteriorly it extends along the ventral margin of the temporal arch to about the same level as the postorbital does along the dorsal margin of the temporal arch. Together, these two elements embrace the anterior tip of the squamosal, which remains restricted to a level behind the anterior margin of the upper temporal fossa in Protenodontosaurus and hence does not reach as far anteriorly as in Placochelys.
The parietal skull table, formed by the fused parietals, is broad and shows straight lateral mar- gins. The large pineal foramen is located anteri- orly in the skull table (Fig. 27B). Dermal encrus- tations are present but poorly defined. As far as can be determined, they resemble those of Pla- cochelys, with a pair of encrustations along either side of the parietal skull table and an unpaired posteromedial encrustation located at the posterior margin of the skull table. The diverging posterior processes of the parietal meet the squamosals along the posteromedial margins of the upper temporal opening in an interdigitating suture, which is distinct in the second specimen (mfsn 1923GP) only. Dermal encrustations create a dis- tinct step at the posterior corner of the upper tem- poral fossa on both sides of the skull that might be mistaken for a suture separating a small squa- mosal from the large quadratojugal. Whether the quadratojugal entered the posterolateral margin of the upper temporal fossa remains indistinct in Protenodontosaurus because of dermal encrusta- tions on the temporal arch, which obscure the su- tural relations, and because of the incompleteness of the specimens.
The ventral view of the skull (Fig. 27C) shows the much enlarged vomers, which extend far for- ward into the rostrum, an autapomorphy of Pro- tenodontosaurus. The enlarged vomers exclude the premaxillaries from the internal nares. Poste- riorly the vomers do not extend beyond the level of the posterior margins of the internal nares as they do in Cyamodus kuhnschnyderi and Placo- chelys (unknown in Cyamodus rostratus and Pse- phoderma). The maxilla narrowly enters the an- terolateral margin of the internal naris, between the anterior vomer and the posterior palatine. A distinct nutritive (dental lamina) foramen is locat- ed within the maxilla behind the maxillary tooth plate. In ventral view, the posterior end of the maxilla tapers to a pointed tip that remains ex-
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
57
Table 10. Measurements of the maxillary and pal- atine tooth plates of Protenodontosaurus italicus (holo- type, mfsn 1819GP). All measurements in mm.
|
left |
right |
|||
|
longi- tudinal 0 |
trans- verse 0 |
longi- tudinal 0 |
trans- verse 0 |
|
|
maxillary tooth |
20.2 |
13.4 |
20.0 |
13.3 |
|
anterior palatine tooth |
12.8 |
13.3 |
13.5 |
13.2 |
|
xjsterior palatine tooth |
36.0 |
28.5 |
36.0 |
29.0 |
eluded from the anterior margin of the subtem- poral fossa by a contact of the jugal and palatine. On the left side of the skull of the holotype (mfsn 1819GP), a lateral process of the palatine meets the jugal, as is also the case in Placochelys; on the right side of the same specimen, the jugal is seen to extend medially to meet the palatine. In Protenodontosaurus, the jugal does not extend backward along the anteromedial margin of the subtemporal fossa, as it does in Cyamodus rostra- tus.
As in other cyamodontoids, the palatine carries two tooth plates, of which the posterior one is much larger than the anterior one (Table 10). The posterior dental lamina foramina are located be- hind the posterior palatine tooth plates on the pal- atine-pterygoid suture. Although the posterior palatine tooth plates are not as disproportionally enlarged as in Macroplacus, the relative palatal exposure of the pterygoid is very short in Proten- odontosaurus. The ratio of the length of the pter- ygoid (from its posterior margin to the dental lam- ina foramen on the palatine-pterygoid suture) to the distance from the pterygoid-palatine suture to the posterior margin of the internal naris is 0.2 for Protenodontosaurus. This indicates that the rela- tive length of the palatal exposure of the ptery- goid is not simply a reflection of the relative size of the posterior palatine tooth plates. The ptery- goid forms a longitudinally oriented ventral flange with a single ventral projection. The pterygoid ex- tends anteriorly along the medial margin of the subtemporal fossa to the level of about the mid- point of the longitudinal diameter of the posterior palatine tooth plate. An ectopterygoid is absent in Protenodontosaurus.
The lateral view of the braincase shows the dis- tinct palatoquadrate cartilage recess with a com- paratively broad contact of the palatine with the anteromedial wing of the quadrate along its lateral edge (Fig. 27B). The posterodorsal margin of the
palatoquadrate cartilage recess is formed by the otic process of the squamosal, which in Proteno- dontosaurus extends far anteriorly, closely ap- proaching the trigeminal incisure located between the prootic and epipterygoid. As mentioned above, the second specimen (mfsn 1923GP) shows an internal subdivision of the trigeminal incisure by a vertical strut of bone. The epitery- goid is a broad element that in the second speci- men (mfsn 1923GP) is incompletely ossified. Un- fortunately, the dorsal relations of the epiptery- goid remain incompletely known for Protenodon- tosaurus. In particular, it is impossible to ascertain whether the epipterygoid forms a posterior dorsal process that meets the squamosal at the dorsal margin of the posttemporal fossa.
The posttemporal fossa is large in Protenodon- tosaurus, with the pteroccipital foramen located at its ventral margin. As in Placochelys, the pter- occipital foramen appears slightly recessed below and behind the ventral margin of the posttemporal fossa (Fig. 27B). The exoccipital forms the lateral margin of the foramen magnum and between it- self and the opisthotic encloses the jugular fora- men. The jugular foramen is not subdivided in- ternally in Protenodontosaurus. The exoccipitals meet each other dorsal to the occipital condyle formed by the basioccipital alone, a character shared with Cyamodus kuhnschnyderi (Nosotti & Pinna, 1996) but absent in Placochelys (and Pse- phoderma: see below). As in other cyamodon- toids, the anterior aspect of the relatively slender paroccipital process is lined by the squamosal, but a distinct squamosal buttress receiving the distal end of the paroccipital process, as seen in Pla- cochelys (and Psephoderma: see below), is absent in Protenodontosaurus. The foramen for the in- ternal carotid is closed ventrally in Protenodon- tosaurus by a ventral contact of the basioccipital tuber and the ventral process of the opisthotic, yet the opisthotic pedicel remains widely separated from the posterior margin of the basicranium or pterygoid, respectively.
Enlarged temporal tubercles secondarily fused to the underlying bone are absent on the lateral surface of the temporal arch of Protenodontosau- rus, whereas they do occur in Cyamodus kuhn- schnyderi and Placochelys. Unfortunately, the posterior margin of the upper temporal fenestra is poorly preserved on both sides of the skull of Protenodontosaurus, so that it is impossible to as- certain whether such temporal tubercles were re- stricted to a posterior position as in Psephoderma or were altogether absent in Protenodontosaurus.
58
FIELDIANA: GEOLOGY
Fig. 28. 20 mm.
Skull of Psephoderma alpinum H. v. Meyer (msnm V471): A, dorsal view; B, ventral view. Scale bar =
The Cranial Anatomy of Psephoderma alpinum H. v. Meyer, 1858
Psephoderma alpinum was originally based on an isolated carapace (H. v. Meyer, 1858) from the Rhaetian Koessen-Formation of the Bavarian Alps (Winkelmoos Alpe), but today it is known from several articulated specimens, two of which are complete (Pinna & Nosotti, 1989; Renesto & Tin- tori, 1995). All of these specimens have sustained severe dorsoventral compression, which has ob- scured some details of skull anatomy. The speci- men yielding the most information on the cranial anatomy of Psephoderma is the isolated skull from Monte Cornizzolo (msnm V471; Pinna, 1976a; Pinna & Nosotti, 1989). This skull too was subjected to some dorsoventral compression, however, which resulted in extensive breakage in the preorbital region. Also, the specimen is in- completely prepared: both orbits and the left tem- poral fossa are still filled with matrix, and the lat-
eral wall of the braincase is only incompletely ex- posed in the right temporal fossa (Fig. 28).
Other than by the morphology of the rostrum (see above), the genus is unique among cyamo- dontoids in the relative proportions of the upper temporal fossa. The skull is depressed yet narrow, and the temporal fossae in particular are relatively short and distinctly narrower than in other cy- amodontoids (Table 8).
The paired premaxillaries form the elongated, narrow, edentulous rostrum (Fig. 29). Posteriorly the premaxillaries extend into distinctly enlarged posterior (nasal) processes that project backward far beyond the level of the posterior margins of the external nares, entering deeply in between the frontals (Pinna & Nosotti, 1989). The premaxillae thus completely separate the nasals from one an- other, a character that Psephoderma (Fig. 29B) shares with Macroplacus. The nasals are reduced to small, elongate elements located entirely pos- terior to the external nares, between the premax-
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
59
qj Pp
60
FIELDIANA: GEOLOGY
Fig. 29. Continued.
ilia and the prefrontal, although their precise out- line can no longer be determined because of breakage. A break runs transversely through the posterior (nasal) processes of the premaxillaries at the level of the anterior margin of the external nares, creating a distinct step that might be mis- taken for a separation of the premaxillaries from large nasals.
The exact contours of the prefrontal are also difficult to identify because of breakage, but, as in all other cyamodontoids, a lacrimal is absent. The maxilla is an elongated yet very low element that extends anteriorly into the rostrum (see above) and posteriorly to a level slightly behind the midpoint of the longitudinal diameter of the orbit, where it meets the jugal in a V-shaped su-
Fig. 29. Skull of Psephoderma alpinum H. v. Meyer (msnm V471 ): A, left lateral view, B, dorsal view; C, ventral view. Scale bar = 20 mm. For a list of abbreviations see p. 3.
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
61
ture (with the apex pointing backward; Fig. 29A). An ascending process of the maxilla cannot be identified, but even if this should be attributable to breakage, the process must have been rudimen- tary, given the low profile of the skull. The pos- terior end of the maxilla extensively enters the anterior ventral (lateral) margin of the orbit. Be- cause of incomplete preparation, the lacrimal fo- ramen cannot be identified. The right side of the skull would seem to indicate that the jugal did not extend anteriorly along the ventral margin of the orbit beyond the level of the midpoint of the lon- gitudinal diameter of the latter, although if true, the maxilla would extend along the medioventral margin of the orbit much farther back than is in- dicated by the lateral exposure of the bone. The left side of the skull shows no indication of the anterior tip of the jugal.
The frontals are paired, each forming a distinct anterolateral process that is embraced by the pre- maxilla and prefrontal (Fig. 29B). A contact of the prefrontal with the nasal separates the frontal from the maxilla. Between prefrontal and post- frontal, the frontal forms the concave dorsal mar- gin of the orbit. The large and elongate pineal foramen lies at the anterior end of the parietal, at a level between the orbit and the upper temporal fossa. On the right side of the skull, a narrow entry of the frontal into the anterior margin of the pineal foramen can be identified. The postfrontal is a comparatively small, triangular element lo- cated at the posterodorsal corner of the orbit. Its posterolateral margin is concave but evenly curved rather than angulated. A contact of the postorbital with the parietal broadly separates the postfrontal from the anteromedial margin of the upper temporal fossa.
The postorbital is a large element that forms most of the broad postorbital arch (Figs. 29A, B). It defines the posteroventral (posterolateral) mar- gin of the orbit. Along the lateral margin of the upper temporal fossa, the postorbital extends backward to a level well behind the midpoint of the longitudinal diameter of the upper temporal fossa, a character that Psephoderma (Fig. 29B) shares with Placochelys and Macroplacus but not with Protenodontosaurus, in which the postorbital extends only to about the midpoint of the longi- tudinal diameter of the upper temporal fossa (this character remains unknown for Cyamodus).
The parietals of Psephoderma are fused and form a broad parietal skull table with distinctly concave lateral margins due to a constriction of
its posterior part. Instead of dermal encrustations, the skull table is ornamented with a pattern of ridges and grooves radiating from its center of ossification to the margins of the bone. Behind the pineal foramen, the concave anteromedial margin of the fused parietal forms a distinct step that con- tinues laterally along the anteromedial margin of the postorbital. In this way the pineal foramen comes to lies in a distinct depression, which is a unique character of Psephoderma. The right side of the skull shows a distinct yet slender anterior process of the parietal that extends beyond this step medial to the postorbital, postfrontal, and frontal to form the lateral margin of the pineal foramen.
The left temporal arch shows particularly well the sutures delineating the posterior processes of the postorbital and jugal, together embracing the anterior process of the squamosal which, unlike in Placochelys, extends anteriorly to a level be- yond the anterior margin of the upper temporal fossa (Fig. 29A). The quadratojugal extends an- teriorly along the ventral margin of the temporal arch beyond the level of the midpoint of the lon- gitudinal diameter of the upper temporal fossa, but it does not reach beyond the level of the an- terior margin of the upper temporal fossa, as is the case in Placochelys.
Because of breakage of the bone surface, the relations of the vomers remain obscure in the ven- tral view of the skull of Psephoderma (Fig. 29C), but, especially on the left side of the skull, it ap- pears that the premaxilla enters the anterior mar- gin of the internal naris, whereas it remains ex- cluded therefrom in all other cyamodontoids, with the possible exception of Macroplacus. The max- illa enters the anterior lateral margin of the inter- nal naris, between the premaxilla anteriorly and the palatine posteriorly. The left side of the skull shows a distinct medial process of the jugal, which meets the palatine at the anteromedial mar- gin of the subtemporal fossa and thus excludes the maxilla from the latter. There is no indication that the jugal extends backward along the medial mar- gin of the subtemporal fossa, as it does in Cy- amodus rostratus.
The posterior palatine tooth plates of Psepho- derma are distinctly elongated (at least in the adult, see Tables 7 and 1 1 ), a character also shared by Macroplacus. The posterior dental lamina fo- ramina are located on the pterygoid-palatine su- ture posteromedial to the posterior palatine tooth plates. In spite of the elongation of the palatine
62
FIELDIANA: GEOLOGY
tooth plates, the pterygoid retains a relatively long palatal exposure compared with Protenodontosau- rus (without elongation of the posterior palatine tooth plates) and Macroplacus (with much en- larged posterior palatine tooth plates). The relative length of the palatal exposure of the pterygoid can therefore be treated as a character that, to some degree at least, is independent from the relative size of the posterior palatine tooth plates (see also the discussion above). Dividing the length of the pterygoid (from its posterior margin to the dental lamina foramen on the palatine-pterygoid suture) by the distance from the pterygoid-palatine suture to the posterior margin of the internal naris yields a ratio of approximately 0.36 for Psephoderma (msnm V471). The pterygoid forms a longitudi- nally oriented ventral flange with a single ventral projection. The left pterygoid is seen to extend anteriorly along the medial margin of the subtem- poral fossa to the level of the posterior third of the longitudinal diameter of the posterior palatine tooth plate. An ectopterygoid is absent (Fig. 29C).
The lateral wall of the braincase is only par- tially exposed, and little anatomical detail is re- vealed because of poor preservation and prepa- ration. The palatoquadrate cartilage recess is dis- tinct, however, and the palatine contacts the an- teromedial lamina of the quadrate along its lateral edge.
The occiput of Psephoderma is deeply exca- vated. The occipital exposure of the braincase was subject to erosion, which obscured structural de- tail. The quadratojugal is clearly demarcated from the quadrate in occipital view. There is a distinct buttress on the lower surface of the left squamo- sal, abutted by the distal tip of the slender par- occipital process (Fig. 29B). A comparable squa- mosal buttress is present in Placochelys, but not in Cyamodus or in Protenodontosaurus. The ven- trally descending flange of the opisthotic is dis- tinct and in Psephoderma contacts the posterior margin of the pterygoid. Among the other cyamo- dontoids included in this study, a comparable con- tact is observed only in the Berlin specimen of Placochelys (mb.r. 1765).
Compared with Cyamodus and Placochelys, en- larged temporal tubercles secondarily fused to the underlying bone are reduced in Psephoderma, where they are restricted to the posterior extremity of the squamosals but do not appear on the lateral surface of the posterior part of the temporal arch (as indicated in the reconstruction of Pinna & No- sotti, 1989).
Autapomorphies in the Skull of the Cyamodontoidea
In addition to highly diagnostic characters of the postcranium, such as the development of an extensive dermal armor or dermal encrustations on the skull, the cranial anatomy offers additional evidence of the monophyly of the Cyamodonto- idea (Rieppel & Zanon, 1997). The palatal den- tition is reduced to two tooth plates on the palatine (one in Henodus: Huene, 1936). The postfrontal remains excluded from the upper temporal fenes- tra by a contact of the postorbital with the parietal (the postfrontal enters the anteromedial margin of the upper temporal fossa in Placodus [Rieppel, 1995a], a condition that is also plesiomorphic for Eosauropterygia [Rieppel, 1997]). The pineal fo- ramen is large, displaced anteriorly, and may bor- der on the frontoparietal suture. The jugal extends far backward in the temporal arch, contacting the quadratojugal and the squamosal posteriorly (per- haps also in Placodus, but not in Paraplacodus [Rieppel, 1995a]; a short jugal was reconstructed for Cyamodus kuhnschnyderi by Nosotti & Pinna, 1996, but the temporal arch is incomplete in both specimens). The ectopterygoid is missing (but see the discussion of Placochelys above). Most dis- tinctive for cyamodontoids, however, is the struc- ture of the secondary lateral wall of the braincase, which incorporates the palatoquadrate. The epi- pterygoid is very broad. It extensively overlaps the dorsal surface of the palatine and is connected to the quadrate by palatoquadrate cartilage, which persisted in the adult. The palatine contacts the quadrate lateral to the palatoquadrate cartilage re- cess in some taxa, and a groove on the dorsal surface of the palatine accommodates the persis- tent anterior (palatine) process of the palatoquad- rate. Posteriorly, the palatoquadrate cartilage re- cess is floored by the pterygoid, the dorsal wing of which broadly overlaps with an anterior wing of the quadrate. Together with the fusion of the palatobasal articulation, this broad overlap of pterygoid and quadrate obliterates the anterior part of the cranioquadrate passage. As a conse- quence, the stapedial (temporal) artery reaches the jaw adductor musculature through the pteroccipi- tal foramen. Finally, the internal carotid enters the cranioquadrate passage through a foramen located between the basioccipital tuber and a ventral pro- cess of the opisthotic and continues anteriorly in a basicranial canal. This whole character complex is unique among amniotes but is shared by all cyamodontoid placodonts.
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
63
Fig. 30. Left lateral braincase wall of Placodus gigas Agassiz (umo BT 13). Scale bar = 20 mm. For abbrevia- tions, see p. 3.
Because the clade that includes Placodus and Paraplacodus is the sister group of the Cyamo- dontoidea, it is interesting to compare the derived structure of the cyamodontoid braincase with the more generalized braincase anatomy of Placodus (the braincase of Paraplacodus is not known). The best-preserved and best-prepared skull of Placodus is the specimen umo BT 13, originally described by Sues (1987) and redescribed by Rieppel (1995a). In contrast to cyamodontoids, Placodus has an epipterygoid with a broad base but a relatively narrow dorsal process. Broili (1912, PI. 14, Fig. 6) described a recess located between the posteroventral aspect of the base of the epipterygoid (covered by unfinished bone) and the anteromedial aspect of the quadrate, above the pterygoid. Huene (1931) correctly assumed that this recess must have housed persistent palato- quadrate cartilage. The specimen umo BT 13 shows nicely (Fig. 30) the extensive dorsomedial flange of the quadrate ramus of the pterygoid, which anteriorly supports the broad, fan-shaped base of the epipterygoid. The dorsal head of the epipterygoid abuts the descending flange of the parietal at the posterodorsal corner of the foramen interorbitale. A bony process rises up along the medial aspect of the anterior margin of the epi-
pterygoid, which is clearly located lateral to the rostrum basisphenoidale and hence cannot repre- sent an ossification of the primary lateral wall of the braincase. Its base is broken in umo BT 13, but Broili's (1912) specimen suggests that this dorsal process originates from the pterygoid. Oth- er than in cyamodontoids, the cranioquadrate pas- sage persists in Placodus; it extends from the space between the pterygoid and the dorsolateral aspect of the basioccipital tubers into the cavum epiptericum medial to the epipterygoid (Rieppel, 1995a). The cavum epiptericum is represented by the space between the epipterygoid and the ros- trum basisphenoidale. The rostrum basisphenoi- dale carries the sella turcica, which is pierced by two foramina for the passage of the cerebral ca- rotids into the braincase. From the lateral margin of the sella turcica originates, on either side of the skull, a clinoid process that represents an ossifi- cation of the primary lateral wall of the braincase (base of the pila antotica). The clinoid process rises up in a dorsolateral direction and contacts the dorsal tip of the dermal process lining the an- teromedial edge of the epipterygoid. The anterior opening of the cavum epiptericum thus is well- defined by the clinoid process and the dorsal pro- cess of the ?pterygoid lining the anteromedial
64
FIELDIANA: GEOLOGY
margin of the epipterygoid. The internal carotid must have entered the cranioquadrate passage be- tween the pterygoid and the basioccipital tube. It must have subdivided within the posterior part of the cranioquadrate passage, giving rise to the sta- pedial artery, which must have reached the tem- poral musculature by passing through a gap be- tween the dorsomedial flange of the quadrate ra- mus of the pterygoid and the descending flange of the parietal, lateral to the otic capsule and an- terior to the paroccipital process (Fig. 30).
As in cyamodontoids, the palatine-pterygoid suture can be traced to the ventral tip of the lon- gitudinally oriented ventral flange of the ptery- goid. From there it trends in an anterodorsal di- rection until it disappears under the epipterygoid. In Placodus, the epipterygoid does encroach on the posterior margin of the palatine, but it does not invade the dorsal surface of the palatine to the same extent as the much broader epipterygoid of cyamodontoids. Posteriorly, the base of the epi- pterygoid expands over the dorsomedial flange of the pterygoid, terminating in a surface of unfin- ished bone. As in cyamodontoids, the quadrate of Placodus also carries a distinct anteromedial flange that overlaps the dorsomedial flange of the pterygoid and again terminates in an unfinished margin opposing the posteroventral margin of the base of the epipterygoid. In between these two bones, the pterygoid flange must have been cov- ered by persistent palatoquadrate cartilage in the live animal (Fig. 30). In contrast to cyamodon- toids, there is no evidence that a cartilaginous an- terior palatal process of the palatoquadrate per- sisted in Placodus.
As mentioned above, the stapedial artery of Placodus passed through a gap between the dorsal flange of the pterygoid and the descending flange of the parietal. The prootic is exposed at the an- teromedial corner of this gap as it emerges from behind the epipterygoid and from below the de- scending flange of the parietal. Posterolateral to the prootic, the opisthotic extends into the par- occipital process, again exposed in lateral view between the pterygoid and the parietal. The gap for the passage of the stapedial artery is closed posteriorly by the squamosal.
In cyamodontoid placodonts, this gap for the passage of the stapedial artery is reduced to a small pteroccipital foramen (Nosotti & Pinna, 1993b). Closure is the consequence of the for- mation of a neomorphic otic process of the squa- mosal, which extends anteriorly along the dorsal margin of the pterygoid flange and meets the
prootic, which expands backward. These two bones (in Cyamodus rostratus, see above), or the otic process of the squamosal alone (in Placo- chelys), form the anterior margin of the pteroc- cipital foramen. In a parallel development, there is an expansion of the squamosal anteromedially along the anterior aspect of the paroccipital pro- cess; this, together with the opisthotic, forms the posterior margin of the pteroccipital foramen.
Evolution of the Rostrum and of the Dentition Within the Cyamodontoidea
The most distinctive characters that differ among the cyamodontoid taxa relate to the struc- ture of the rostrum and the dentition. The genus Cyamodus retains a short, broad rostrum formed by tooth-bearing premaxillae. There are two teeth in each premaxilla of Cyamodus rostratus (Fig. 12B), and between one (Nosotti & Pinna, 1996) and two (Rieppel & Hagdorn, 1999) teeth on the premaxilla of Cyamodus kuhnschnyderi. Ontoge- netic variations of the dental formula of Cyamo- dus hildegardis have been described by Kuhn- Schnyder (1959).
The rostrum of Macroplacus is incomplete, but what is preserved shows a deep trough, or con- cavity, on its ventral surface between the two pre- maxillaries (Fig. 26C). The maxilla carries an an- terolateral, relatively short yet distinct process that tapers off along the lateral margin of the rostrum well in front of the level of the anterior margin of the internal nares. This sutural pattern contrasts with that of Cyamodus, where the suture between the premaxilla and maxilla is more transversely oriented (C. kuhnschnyderi) or even trends in a posterolateral direction (C. rostratus). The ante- rior tip of the premaxilla does not extend anteri- orly beyond the level of the anterior margin of the internal nares in Cyamodus.
The rostrum of Protenodontosaurus is more elongated than that of Cyamodus, and each pre- maxilla carries an anterior tooth, plus a second posterior alveolus, which may have carried a sec- ond premaxillary tooth or which may represent a replacement pit (Nosotti & Pinna, 1999). As in Macroplacus, the maxilla of Protenodontosaurus forms an anterior process that participates in the formation of the rostrum. However, this antero- lateral process of the maxilla is somewhat longer in Protenodontosaurus, and its ventral surface shows a low ridge that, together with its counter- part from the other side of the skull, delineates a
RIEPPEL: THE CRANIAL ANATOMY OF PLACOCHELYS PLACODONTA
65
Table 1 1 . Data matrix for the analysis of placodont interrelationships. For further details, see text.
|
1 |
1 |
2 |
3 |
4 |
s |
6 |
7 |
8 |
9 |
1 0 |
1 1 |
1 2 |
1 3 |
1 4 |
1 5 |
1 6 |
1 7 |
1 8 |
1 9 |
2 0 |
2 1 |
2 2 |
2 3 |
|
|
1 |
Ancestor |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
2 |
Placodus |
1 |
0 |
0 |
0 |
0 |