Memorie della Società italiana di scienze naturali e del Museo civico di storia naturale di Milano

SECEZIOR:

MEMORIE

della Società Italiana

di Scienze Naturali

e del Museo Civico

di Storia Naturale di Milano

Volume XXXVII - Fascicolo I

É Pa x
CRISTIANO DAL SASSO & CA A
SIMONE MAGANUCO Va,
e La
ì n.

SCIPIONYX SAMNITICUS — @ —_©
(THERQPODA: COMP: PSOGNATHIDAE)
‘HÉ RCRETACEOUS

MOZ

4 nie ©
LIBRARY
MIC 4% 2044
AUG 15 2011

LIADIN/AFCSTX
MAriVArL)

UNIVERSITY

MILANO MAGGIO 2011

54:
pra
ni. |
prîe
pro
ca
pi
piaq
ptect
A
CI
bi

Elenco delle Memorie della Società Italiana di Scienze Naturali e
del Museo Civico di Storia Naturale di Milano —

V- NOT E EEE STA 28 pp.,
3 figg., 2 tavv. EER i

I- MARTORELLI G., 1895 ri agita
in Italia. 2/6.pp., 46figg., A tav. Aia

RELLI G., - Le forme e l di
piumaggio. Memoria ornitologica. //2 pp., ‘ SE
PAVESI P., 1901- L’abbate a iaia 14 figg., 1 tav.
Volume VIT :
I- DE ALESSANDRI G., 1910 - Studi sui pesci triasici della Lombardia.
164 pp., 9 tavv.

I- REPOSSI E., 1915 - La bassa Valle della Mera. Studi petrografici e

ì . Parte I. pp. 1-46, 5 figg., 3 tavv. i
- REPOSSI Bo VOTE TOI aa Valle della Mera. Studi petrogra-
fici e geologici. Parte II. pp. 47-186, 5 figg. 9 tavv.

I-

PINNA G., 1969 - Revisione delle ammoniti figurate da Giuseppe Me-

rieghini nelle Tavy11-20./della
7ouge cmmonitigue» (1867-1881), pp 14002

molari d’elefante delle alluvioni lombarde,
filogenia e scomparsa di alcuni ;

., 6 tavv.

Cristiano Dal Sasso & Simone Maganuco

Scipionyx samniticus (Theropoda: Compsognathidae)
from the Lower Cretaceous of Italy
Osteology, ontogenetic assessment, phylogeny,
soft tissue anatomy, taphonomy and palaeobiology

Volume XXXVII - Fascicolo I

Maggio 2011

Memorie della Società Italiana di Scienze Naturali
e del Museo Civico di Storia Naturale di Milano

CREDITI DELLE ILLUSTRAZIONI

La maggior parte delle immagini pubblicate in questa
monografia è tutelata da diritti d’autore pertinenti ai fotografi,
disegnatori e artisti che le hanno realizzate, e/o alle istituzioni
cui appartengono i soggetti illustrati. Tali immagini, pertanto,
non possono essere utilizzate a fini commerciali senza previo
permesso scritto dei titolari dei diritti d’autore.

Date le limitazioni di spazio imposte dalle didascalie
bilingui, i nomi degli autori delle foto e dei disegni sono
qui elencati in ordine alfabetico, seguiti dai numeri delle
figure; i nomi de gli autori delle ricostruzioni artistiche
sono elencati separatamente, preceduti dai numeri delle
pagine.

Roberto Appiani, © Soprintendenza per i Beni Archeologici di Salerno,
Avellino, Benevento e Caserta: 9, 19B, 23, 28A, 31A, 35A, 40A, 43A,
49:55, 87,97, 103-116,-125; 126,127;0128, 129B, 136B, 137,145,
146B, 159, 164A.

Roberto Appiani & Leonardo Vitola, © Soprintendenza per i Beni
Archeologici di Salerno, Avellino, Benevento e Caserta: 10, 21, 115,
156.

Marco Auditore, © Museo di Storia Naturale di Milano: 5, 8A, 10, 22,
24,25A, 25B, 26, 29, 33, 36B, 37, 38, 41, 50, 53, 56, 60, 62, 64, 67, 68,
787088188 SNO SM 0024717387487 817691735
182, 186.

Armando Cioffi, © Soprintendenza per i Beni Archeologici di Salerno,
Avellino, Benevento e Caserta / Fondazione Ospedale Maggiore di
Milano: 32, 92, 101, 148.

Cristiano Dal Sasso: 2, 7B, 20.

Cristiano Dal Sasso, © Soprintendenza per i Beni Archeologici di Sa-
lerno, Avellino, Benevento e Caserta / Museo di Paleontologia dell’ Uni-
versità di Napoli “Federico II°: 15.

Cristiano Dal Sasso & Simone Maganuco: 34.

Cristiano Dal Sasso & Simone Maganuco, © Soprintendenza per i Beni
Archeologici di Salerno, Avellino, Benevento e Caserta: 8B, 8D, 8F,
8G, 11, 27, 28B, 30, 31B, 31C, 35B, 36A, 39, 40B, 42, 43B, 45, 46,
48,917 52,54;197, 58; 99; 01, 69, 6094107120405 TI16, 19332083,
84, 86, 89, 90, 93, 95, 96, 99, 100, 102B, 104, 105, 106, 107, 108, 122,
123, 124, 130, 131, 133, 134, 135, 136A, 138A, 138B, 140, 142, 147,
150A, 152, 154, 155, 158, 160, 162, 163, 164B, 166, 167, 168, 172,
177, 178B, 179, 180, 184.

© 2011 Società Italiana di Scienze Naturali
Museo Civico di Storia Naturale di Milano
Corso Venezia, 55 - 20121 Milano

ILLUSTRATION CREDITS

Most images used in this publication are copyrighted
by the authors of the photographs, drawings, or
artworks, and/or by the institutions which the illustrated
subjects belong to. Therefore, they may not be used for
commercial purposes without prior written permission
of the copyright holders.

Due to space limitations for the bilingual figure captions,
the names of the authors of photographs and drawings
are listed in alphabetical order, followed by the figure
numbers; the names of the authors of the artworks are
listed separately, preceded by the page numbers.

Fritz Huchzermeyer: 110.

Simone Maganuco: 85, 113.

W. Scott Persons, IV: 160C.

Stefano Scali: 1.

Giorgio Teruzzi: 6, 18, 19A, 150B, 151, 183.

Leonardo Vitola, © Soprintendenza per i Beni Archeologici di Salerno,
Avellino, Benevento e Caserta: 8C, 8E, 8H, 12, 14B, 44, 47, 63, 66, 71,
80, 91, 94, 102A, 129A, 138C, 141, 146A, 153, 161, 165, 185A.

Leonardo Vitola, © Soprintendenza per i Beni Archeologici di Salerno,
Avellino, Benevento e Caserta / Museo di Paleontologia dell’ Università
di Napoli “Federico II”: 14 A, 14C, 14D.

Michele Zilioli, © Soprintendenza per i Beni Archeologici di Salerno,
Avellino, Benevento e Caserta / Museo di Storia Naturale di Milano:
13, 118, 119, 120, 121, 132, 139, 143; 144, 149,1S7,169, 1702008
181, 185B, 185C.

269: © Marco Auditore & Arianna Nicora
270: © Davide Bonadonna

271: © Paolo Cinquemani

272: © Fabio Pastori

273: © Renzo Zanetti

274: © Loana Riboli

275: © Fabio Fogliazza

276: © Lukas Panzarin

277: © Tullio Perentin

278: © Troco

In copertina: Scipionyx samniticus in grandezza naturale, nella porzione di Plattenkalk autentico.
Life-size Scipionyx samniticus, and the authentic portion of Plattenkalk embedding it.

Registrato al Tribunale di Milano al n. 6694
Direttore responsabile : Anna Alessandrello
Grafica editoriale: Michela Mura

Stampa: Litografia Solari, Peschiera Borromeo - Maggio 2011

ISSN 0376-2726

Quando lanatura viene alla generatiò delle/pietre essa genera una qualita
domore visscioso/il quale col suo secharsi congele inse co chede/tro allui
sirinchiude enòli converte inpietra/ma li coserua dentro asse nella forma
che elli ha trovati...

Leonardo da Vinci

Quando la natura crea le pietre genera una specie di umore vischioso che
asciugandosi congela ciò che racchiude in se stesso e non trasforma tutto in pietre
ma conserva ciò che racchiude nella forma trovata...

When nature creates stones she generates a type of viscous humour that upon
drying freezes within it whatever is found, not transforming those things into
stone but conserving them in the form with which they were found....

Cristiano Dal Sasso & Simone Maganuco

Scipionyx samniticus (Theropoda: Compsognathidae)
from the Lower Cretaceous of Italy
Osteology, ontogenetic assessment, phylogeny,
soft tissue anatomy, taphonomy and palaeobiology

Abstract — This monograph provides a detailed account of the remarkably preserved holotype of the theropod dinosaur Scipionyx
samniticus uncovered in the Lower Cretaceous (Albian) of Pietraroja, southern Italy. We describe the geological setting of the only
known specimen and give a short description of the lithology and stratigraphy of this famous, fossil-rich locality. Macroscopic and
microscopic observations, as well as chemical analysis of the sediment embedding the fossil, has allowed better estimation of its posi-
tion within the stratigraphic series of the outcrop. An overview of the depositional environments hypothesised by previous authors for
the Pietraroja fossils is presented, although the taphonomical study of Scipionyx suggests that this specimen was buried in a single,
rapid event by a turbidite. We also include palaeoenvironmental and palaeogeographic data commented in the light of the Cretaceous
dinosaur record of the Periadriatic region, focusing in particular on the central and southern Italian tracksites and their palaeobiogeo-
graphical significance.

Part I is dedicated to the osteology of the specimen. The major cranial novelties described include the identification of formerly
unrecognised braincase bones, and reinterpretation of palatal and mandibular elements, which lead to better understanding of their
topology. Other relevant cranial features of the specimen are 5 premaxillary teeth, a sinusoidal ridge of the supratemporal fossa with
frontoparietal contact, a distally squared descending process of the squamosal, lower tooth row extending farther back than the upper
row, and the absence of an external mandibular fenestra. Among the relevant postcranial skeletal features described are fan-shaped
dorsal neural spines with beak-like ligament attachments, hair-like cervical ribs, dorsal ribs with cup-like sternal attachments, gastralia
with peculiar morphology and ?pathology, carpus composed of only two stacked, well-ossified bones, manual digit III longer than digit
I, a cranially notched iliac preacetabular blade and a distally squared ischial obturator process.

We then expound on the immaturity of the specimen — which was probably less than three weeks old at the time of death — as sug-
gested by a long list of juvenile characters, such as the presence of a frontoparietal fontanelle, a short and deep antorbital region, tooth
replacement not yet started, peculiar scarred bone surfaces, non-sutured girdle elements and closure of the neurocentral sutures not yet
started in any vertebra. At the end of this part, we give a phylogenetic analysis of Coelurosauria (90 taxa, 360 characters), evaluating
also the ontogeny-related characters that identify Scipionyx as a basal member of a monophyletic Compsognathidae, which results to
be more derived than Tyrannosauroidea.

Part II is on the superbly preserved internal organs and soft tissues of Scipionyx samniticus. The preserved internal tissues include
axial ligaments, axial and appendicular articular cartilage, neck muscles and connective tissue, part of the trachea, oesophageal
remains, traces of the liver and other blood-rich organs, the entire intestine, mesenteric blood vessels and pelvic and hindlimb muscles.
External soft tissues are beautifully represented by the horny manual claws. Most of the soft tissues can be easily identified by their
ochre colour, whereas other organic remains are preserved as thin films that are visible under ultraviolet-induced fluorescence (UV).
Scanning electron microscopy (SEM) analyses have revealed an exceptional three-dimensional preservation of the soft tissues and
astonishing information at a cellular and even subcellular level, such as the sarcomere-related banded pattern observable within every
single muscular myofibre. SEM element microanalysis has also confirmed the haematic origin of the reddish macula formerly referred
to the liver. On the other hand, it has been established that the remains or imprints purported by some authors to be of the diaphrag-
matic muscles are, in fact, a calcite nodule of amorphous microstructure, inconsistent with the preservation of other muscle tissue in
this specimen. This evidence, and other anatomical observations on bones and soft tissues, deny the hypothesis of an hepatic-piston
assisted breathing mechanism in Scipionyx. The exceptional preservation of labile soft tissue indicates that, after death, the carcass of
this theropod hatchling was subjected to very little decay and rapid authigenic mineralisation in the presence of a high concentration
of phosphates. We examine this process in the taphonomy section of this monograph.

Part III focuses on the functional morphology and the palaeobiology of Scipionyx samniticus. Outstandingly, the degree of pre-
servation of the soft tissues has permitted an analysis of the relative position of the food remains in the digestive apparatus and, thus,
reconstruction of a feeding chronology for this specimen, an insight that is usually impossible to obtain for fossil vertebrates. In fact,
Scipionyx’s gut has been found to contain allogenous bones from a lepidosaurian reptile in the stomach region, lizard-like polygonal
squamae in the duodenum, fish scales in the rectum, and a variety of tiny remains in several points of the intestine. This is compelling
evidence that Scipionyx fed on both lizards and fish. Finally, remarks on the digestive physiology and respiratory physiology of Scipio-
nyx, inferred from the internal organs and their osteological correlates, close the monograph.

The amount and detail of information gained from this single specimen make the Pietraroja Plattenkalk a unique fossil locality.
In contrast to the Chinese Jehol Group, which is a lacustrine/volcanic freshwater deposit that has preserved dinosaurs with delicate
integumentary structures, such as filaments, feathers and bristles, the Italian shallow marine Lagerstàtte has preserved internal organs.
This is unprecedented not only for a dinosaur, but also for any other Mesozoic terrestrial vertebrate.

Key words: Theropoda, Compsognathidae, soft tissue preservation, Lower Cretaceous, Pietraroja Plattenkalk, Italy.

Riassunto — Questa memoria monografica descrive in dettaglio il dinosauro teropode Scipionyx samniticus, proveniente dal Cre-
taceo inferiore (Albiano) di Pietraroja (Benevento). L’inquadramento geologico dell’olotipo, che è straordinariamente ben conservato
allo stato fossile e rappresenta tuttora l’unico esemplare conosciuto di questa specie, è qui richiamato con una breve trattazione della
litologia e della stratigrafia della celebre località fossilifera. Lo studio del sedimento in cui giace il fossile, condotto a livello macro-
scopico e microscopico e comprendente anche l’analisi della sua composizione chimica, ha permesso di identificare con buona appros-

CRISTIANO DAL SASSO & SIMONE MAGANUCO

simazione la posizione dell’esemplare all’interno della serie stratigrafica del giacimento. Sebbene lo studio tafonomico di Scipionyx
suggerisca che questo esemplare sia stato sepolto da una torbidite in un singolo e rapido evento, viene presentata una panoramica dei
possibili ambienti di deposizione ipotizzati da altri autori per i fossili di Pietraroja. Vengono qui portati anche dati paleoambientali e
paleogeografici, commentati alla luce della documentazione fossile riguardante i dinosauri nel Cretaceo della regione periadriatica, con
particolare attenzione ai giacimenti di impronte dell’Italia centro-meridionale, e alla loro importanza paleobiogeografica.

L’intera osteologia dell’esemplare viene qui revisionata e descritta osso per osso (Parte I). Nel cranio, le più importanti novità
consistono nell’identificazione di un maggior numero di ossa della scatola cranica; gli elementi del palato e della mandibola sono
stati reinterpretati e la loro topologia è ora più chiara. Le caratteristiche più importanti del cranio sono: 5 denti premascellari, una
cresta sinusoidale presente sul margine della fossa sopratemporale a livello del contatto frontoparietale, un processo discendente dello
squamoso che termina con una estremità squadrata, la fila dei denti inferiori che si estende in direzione caudale più della fila dei denti
superiori, l’assenza della finestra mandibolare esterna. Rilevanti caratteri scheletrici postcraniali sono: le spine neurali delle vertebre
dorsali espanse a forma di ventaglio e con attacchi dei legamenti a forma di becco, le costole cervicali filiformi, gli attacchi delle
costole sternali, la morfologia dei gastralia e la condizione ?patologica di alcuni di essi, il carpo composto soltanto da due ossa ben
ossificate e impilate una sull’altra, il terzo dito della mano più lungo del primo, la lama preacetabolare dell’ileo che porta una incisura
craniale, il processo otturatore dell’ischio squadrato in direzione distale.

L’immaturità dell'esemplare, che al momento della morte aveva probabilmente meno di tre settimane, è confermata da una lunga
lista di caratteri giovanili, come la presenza della fontanella frontoparietale, la regione antorbitale corta, la sostituzione dei denti non
ancora iniziata, la particolare tessitura solcata delle ossa, gli elementi dei cinti non suturati e la chiusura della sutura neurocentrale non
ancora iniziata in alcuna vertebra.

L’analisi filogenetica dei Coelurosauria (360 caratteri analizzati in 90 taxa), in cui si è tenuto conto anche dei caratteri legati
all’ontogenesi, pone Scipionyx come membro basale dei Compsognathidae, che risultano un clade monofiletico e più derivato dei
Tyrannosauroidea.

Una sezione specifica di questa memoria (Parte II) è riservata alla descrizione degli organi interni e di diversi altri tessuti molli,
che nell’olotipo di Scipionyx samniticus sono conservati in modo incomparabile. Tra i tessuti interni che sono fossilizzati vi sono
legamenti intervertebrali, cartilagini articolari dello scheletro assiale e appendicolare, muscoli e connettivi del collo, parte della tra-
chea, residui dell’esofago, tracce del fegato e di altri organi ricchi di sangue, l’intero intestino, vasi sanguigni mesenterici, muscoli del
cinto pelvico, degli arti posteriori e della coda. I tessuti esterni sono superbamente rappresentati dagli artigli cornei, ancora presenti
sulle ultime falangi delle dita delle mani. In gran parte i tessuti molli conservati nell’olotipo di Scipionyx samniticus sono visibili ad
occhio nudo, grazie al colore ocra che ben li distingue dal bruno scuro delle ossa. Altri resti organici sono conservati sotto forma di
sottili pellicole, che possono essere viste solo in fluorescenza indotta da luce ultravioletta (UV). Con esami dettagliati al microscopio
elettronico a scansione (SEM), il nostro studio dimostra che in questo fossile risalente a 110 milioni di anni fa i tessuti molli non sono
semplici impronte, ma sono mineralizzati in tre dimensioni, con una fedeltà di conservazione eccezionale, che raggiunge livelli cellu-
lari e talvolta subcellulari (per esempio, all’interno di ogni singola miofibra muscolare è conservata la struttura a bande dei sarcomeri).
La microanalisi degli elementi al SEM ha confermato l’origine ematica della macchia rossastra precedentemente attribuita al fegato.
Per contro, i residui o impronte di muscoli diaframmatici presunti da alcuni autori in realtà appartengono ad un nodulo di calcite, che
a livello microscopico mostra una struttura amorfa, non compatibile con la conservazione degli altri tessuti muscolari di Scipionyx.
Questa evidenza, unita ad altre osservazioni anatomiche sulle ossa e sui tessuti molli, smentisce l’ipotesi di una meccanica respiratoria
assistita da pistone epatico (una soluzione fisiologica che caratterizza i coccodrilli odierni).

La conservazione eccezionale di tessuti molli molto delicati indica che, dopo la morte, la carcassa del teropode neonato subì una
decomposizione molto limitata e una rapida mineralizzazione autigena, in presenza di una alta concentrazione di fosfati. Questo pro-
cesso è esaminato in una sezione dedicata alla tafonomia dell’esemplare.

Questa monografia indaga anche sulla morfologia funzionale e sulla paleobiologia di Scipionyx samniticus (Parte III), grazie a resti
di cibo che nei primi esami dell’esemplare non erano stati notati. La conservazione dei tessuti molli consente di ricostruire perfino una
cronologia della dieta del piccolo dinosauro, attraverso la posizione relativa dei resti allogeni contenuti nel suo apparato digerente: un
dato impossibile da ricavare per quasi tutti gli organismi fossili. I visceri di Scipionyx contengono: nella regione dello stomaco, le ossa
della caviglia di un rettile lepidosauro; nel duodeno, lembi di pelle composta da squame poligonali riferibili ad un rettile lacertiforme;
nel digiuno, un ammasso di vertebre che probabilmente appartengono ad un piccolo pesce; nel retto, scaglie di un pesce più grande;
in diversi punti dell’intestino, una varietà di altri resti più minuti. Pertanto abbiamo una prova evidente che Scipionyx si nutriva sia di
piccoli rettili sia di pesci, e che questi ultimi rappresentavano un cibo abituale.

La monografia si chiude con alcune considerazioni sulla fisiologia digestiva e respiratoria di Scipionyx, suggerite dallo studio degli
organi interni e dei loro correlati osteologici. La quantità di informazioni e i dettagli che queste hanno fornito, grazie allo studio di un
singolo esemplare, fanno del Plattenkalk di Pietraroja un sito paleontologico unico al mondo. A differenza dei giacimenti cinesi appar-
tenenti al Jehol Group (depositi di origine lacustre/vulcanica che conservano dinosauri fossilizzatisi con strutture tegumentarie delicate
quali filamenti, penne e setole), il Lagerstitte marino italiano ha permesso la conservazione di tessuti molli all’interno del corpo di un
organismo fossile mai visti in precedenza, non solo in un dinosauro, ma anche in qualsiasi altro vertebrato terrestre mesozoico.

Parole chiave: Theropoda, Compsognathidae, conservazione dei tessuti molli, Cretaceo inferiore, Plattenkalk di Pietraroja,
Italia.

INTRODUCTION

Scipionyx samniticus abruptly entered the limelight
when, on the occasion of its formal naming, it made the
cover of Nature (Dal Sasso & Signore, 1998a). Since
then, it has attracted not only the interest of vertebrate
palaeontologists but also popular imagination throughout
the world. In fact, the Scipionyx fossil is a striking, ar-
ticulated, juvenile coelurosaur with a unique combination
of osteological characters and superbly fossilised soft tis-
sues. The latter render Scipionyx one of the best preserved
dinosaurs known, and a unique specimen within the fossil
record of Mesozoic vertebrates.

Scipionyx samniticus was the first dinosaur fossil body
unearthed in Italy; thus, its discovery was a major event
in the history of Italian palaeontology. In fact, Italy was
considered to be devoid of dinosaur remains until this
conviction was upturned by a striking series of finds con-
sisting mainly of fossilised prints (for longer accounts,
see Dal Sasso, 2001, 2003, 2004). Following a first find in
the Dolomites (Mietto, 1988), more dinosaur trackways
were discovered later at Lavini di Marco, in the Veneto
region of northern Italy (Leonardi & Lanzinger, 1992).
Thereafter, a number of Triassic (Norian) and Jurassic

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 7

(Hettangian to Pliensbachian) tracksites have been found
in the mid-eastern Alps (for a complete list, see Leonardi
& Mietto, 2000; Nicosia et a/., 2005; D’Orazi Porchetti
et al., 2008; Leonardi, 2008). Moreover, Cretaceous (Ap-
tian to Santonian) dinosaur tracks have been uncovered
in central and southern Italy, mainly in the Latium and
Apulia regions (Andreassi ef a/., 1999; Gianolla et al.
2000a, 2000b, 2001; Nicosia ef al. 20004, 2000b, 2007;
Conti et al., 2005; Sacchi et al., 2006, 2009; Petti et al.
20082, 2008b).

In addition to Scipionyx, the skeletal remains of three
kinds of non-avian dinosaurs have been found so far
in Italy. Two of them belong to new, possibly endemic,
taxa that inhabited regions of the Periadriatic Domain
during Jurassic (Sinemurian) and Cretaceous (Campa-
nian) times: one taxon is represented by fragmentary
bones of a large theropod that were collected in a com-
mercial quarry in Saltrio (Lombardy region, northern
Italy) — this specimen, still not named officially, has
been described only preliminarily (Dal Sasso, 2001,
2003, 2004); the second taxon is represented by six
specimens, one of which almost complete, of the hadro-
sauroid Tethyshadros insularis, unearthed at Villaggio
del Pescatore, near the town of Trieste, in north-eastern
Italy (Dalla Vecchia, 2009). The presence of a third kind
of dinosaur is documented by a single bone found en-
cased in the eroded wall of a cave from the Cretaceous
(Cenomanian) of Capaci (Palermo, Sicily): recently ex-
amined in situ in cross-section, this bone is the midshaft
of a leg from a medium-large theropod (Garilli et al.,
2009). Given the continuous and frequent rate of these
recent finds, the dinosaur record in Italy is expected to
increase rapidly in the near future.

Uniquely, the footprints and skeletal remains of Ital-
ian dinosaurs were preserved in coastal marine depos-
its. These finds indicate a peculiar palaeogeography and
demonstrate that the model of Bahama-like small islands
proposed for the region is no longer consistent with the
presence of medium-large dinosaurs, which could survive
only in terrestrial ecosystems. As documented by the wide
temporal range of the dinosaur-bearing Italian outerops
(Avanzini et al., 2000; Dalla Vecchia, 2001; Dal Sasso,
2001, 2002, 2003, 2004; Nicosia et a/., 2007; Sacchi et al.,
2009), Mesozoic Italy might have, in actual fact, emerged
several times and quite extensively (see Palaeogeography,
this volume).

The story of the discovery of the Scipionyx fossil is
original in itself (for a detailed account, see Dal Sasso,

2001, 2004). In the spring of 1981, Giovanni Todesco
unearthed a fossil reptile that was rather unusual for the
marine outcrop of Pietraroja (Province of Benevento,
southern Italy). The collector cleaned the fossil as well
as he could and stored it in the basement of his house.
It remained there until 1993, when Todesco showed it to
professional palaeontologists, who identified the tiny rep-
tile, as a dinosaur. The popular magazine Oggi dubbed
the dinosaur “Ciro” (a typical Neapolitan name) in an ar-
ticle they published on it. In accordance with Italian law,
the specimen was handed over to the superintendence of
the site of provenance 1.e., the Soprintendenza per i Beni
Archeologici di Salerno, Avellino, Benevento e Caserta,
where it is still housed. The specimen was first presented
in a brief note (Leonardi & Teruzzi, 1993) and later in a
current publication (Leonardi & Avanzini, 1994). Subse-
quently, a rough examination was conducted, in that the
fossil had still not been fully prepared, and the results de-
scribed in an unpublished degree thesis (Signore, 1995).
In 1994, the Museo di Storia Naturale di Milano (MSNM)
obtained official permission to properly prepare and study
the fossil. It was only during this preparation work, per-
formed in Salerno by Sergio Rampinelli (a collaborator at
the palaeontology laboratory of the MSNM) and one of
the authors of this monograph (CDS), that the extraordi-
nary degree of preservation of the specimen’s soft tissue
was fully realised. This later became the main focus of the
paper published in Nature on its formal description (Dal
Sasso & Signore, 1998a).

The present monograph, concisely introduced by Dal
Sasso & Maganuco (2009), is the most complete publica-
tion existing to date on Scipionyx samniticus. Extensive
re-examination of the specimen was possible thanks to
the Soprintendenza per i Beni Archeologici di Salerno,
Avellino, Benevento e Caserta (SBA-SA), which placed
the fossil on loan to the MSNM from December 2005 to
October 2008. The research laboratories at the MSNM,
and those of collaborators, provided technical support
that was appropriate to the importance of the specimen.
In particular, SEM analysis carried out at the MSNM
provided remarkable evidence of the exceptional level
of the soft tissue preservation. Given that no other sin-
gle dinosaur fossil is known to preserve such a quantity
of previously unknown anatomical features, this mono-
graph is duly devoted to a detailed description of the
unique osteology, ontogenetic stage, phylogeny, soft tis-
sue anatomy, taphonomical history and palaeobiology of
Scipionyx samniticus.

GEOLOGICAL SETTING

The locality

The village of Pietraroja (Benevento Province) sits
near the top of a 970 m high carbonate relief, called “Civi-
ta di Pietraroja”, that rises almost vertically from the plain
adjacent to the eastern margin of the Matese Mountains
(central southern Italian Apennines) roughly 70 km north-
east of Naples (Figs. 1-2). The Lower Cretaceous levels,
assigned to the Lower Albian on the basis of foraminiferal
biozonology (Bravi & Garassino, 1998; Carannante et al.,
2006), crop out at the “Le Cavere” locality, just above the
village 41°20’52.18”°N, 14°32’53.33”E. (The geographi-

cal coordinates 45°77°43.1”°N, 24°82’22.8”’E reported
in recent papers on Pietraroja fossils [Evans e? a/., 2004;
2006] are markedly wrong, as they refer to a locality in
central Romania).

Excavations in the fine-grained marine limestones
have been in progress, albeit intermittently, for more
than 150 years. During this time, the site has yielded
a rich assemblage of plants, invertebrates and verte-
brates, including fishes, amphibians and reptiles (e.g.
Bravi, 1994, 1999; Bravi & Garassino, 1998; Barbera
& Macuglia, 1991; Dal Sasso & Signore, 1998a; Evans
et al., 2006).

CRISTIANO DAL SASSO & SIMONE MAGANUCO

Pietraroja Admin.
[al Boundaries

ci Benevento Da Elevation (meters)

Ma 25-31
ME 832-123
E 129-173
IE 174-216
I 217-259
IMI 260 - 303
BI 304 - 347
MM 348 - 390
MU 391 - 434
MI 435 - 478
uu 479 - 522
BN 523 - 566
I 567-611
um 612- 656
BS 657 - 700
Dai 701-743
Mb 744 - 785
MA 786 - 828
Mi 829 - 872
BE 373-920
N 921-974
BI 975 - 1030
BI 1031 - 1083
BB 1084 - 1136
E 1137 - 1193
BB 1194 - 1260
Bi 1261-1342
BUS 1343 - 1446

1447 - 1594
1595 - 1796

Pietraroja

Campania

9 le) 9 18. Kilometers

Fig. 1 - Geographic location of Pietraroja. A) map of Italy; B) map of the Campania region and its provinces; C) physical map of the
Benevento province and the municipality of Pietraroja.

Fig. 1 - Posizione geografica di Pietraroja. A) mappa dell’Italia; B) mappa della regione Campania e delle sue province; C) mappa
fisica della provincia di Benevento e del comune di Pietraroja.

Fig. 2 - The north-western slope of the “Civita di Pietraroja” and the outerop of “Le Cavere” (centre right). Part of the fossiliferous site
was fenced off after the discovery of the little dinosaur (arrow).

Fig. 2 - La scarpata nord-ovest della “Civita di Pietraroja” e l’affioramento de “Le Cavere” (a destra). Parte del sito fossilifero è stata
cintata dopo la scoperta del piccolo dinosauro (freccia).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 9

Historical background

The limestones of Pietraroja have been known for
their beautifully preserved fossils since the nineteenth
century. The palaeontological richness of the outerop was
mentioned for the first time in 1798 by Breislak (Fig. 3),
but scientific studies began only much later and focused
immediately upon the plentiful number of fossil fishes,
called “ittioliti” (Italian for ichthyoliths), which emerged
from the calcareous rocks. Between 1851 and 1866, Costa

(1851, 1853-1864, 1865, 1866) described the “Calcari ad
Ittioliti di Pietraroja” in a series of volumes superbly il-
lustrated with lithographic prints in the typical style of the
period (Fig. 4). The term “Calcari ad Ittioliti di Pietraroja”
was slightly modified into “calcari selciferi ed ittiolitiferi
di Pietraroja” (Catenacci & Manfredini, 1963), which has
been used frequently to indicate that geological forma-
tion. A formal name has not been validated, yet (Petti,
pers. comm., 2010). Here we use the term Plattenkalk
sensu Carannante ef al. (2006).

Ba iste ar, wr.

LS

va” rei

Scala delle miglia
si i

Fig. 3 - “Physical topography” of north-western Campania, represented in a three-dimensional style. Pietraroja (circle), which is
located some 70 km NE of Naples, is labelled here with the ancient toponym Pietra Roja. (After Breislak, 1798).

Fig. 3 - “Topografia fisica” della Campania nord-occidentale, rappresentata con un aspetto tridimensionale. Pietraroja (area cerchiata),
che si trova circa 70 km a nord-est di Napoli, è qui nominata con l’antico toponimo Pietra Roja. (Da Breislak, 1798).

10

CRISTIANO DAL SASSO & SIMONE MAGANUCO

Law. LL

Carta dir

4 Sa e nti sio

Fig. 4 - Lithography illustrating Belonostomus crassirostris and other fossil fishes unearthed from the Pietraroja Plattenkalk in the

nineteenth century. (After Costa, 1853-1864).

Fig. 4 - Litografia del diciannovesimo secolo che illustra Belonostomus crassirostris e altri pesci fossili del Plattenkalk di Pietraroja.

(Da Costa, 1853-1864).

The age of these limestones — considered to be
Jurassic — was a subject of debate up to the end of
nineteenth century, when Bassani (1885) re-examined
the fossil fishes collected by Costa and confirmed
their Cretaceous affinity. The deposit was studied in
the 1900°s by D’Erasmo (1914, 1915) and D’Argenio
(1963), but generally there was less interest than in the
previous century. In more recent times, research has
been carried out by the Università di Napoli “Federico
II”, which in 1982, in cooperation with the Museo di
Scienze Naturali di Torino, conducted excavations in
the Le Cavere locality (Bravi, 1987, 1988, 1994). Af-
ter the Scipionyx fossil came to light, a small quarry
was opened in 2001 by the MSNM, under the aegis of
the Soprintendenza per i Beni Archeologici di Salerno,
Avellino, Benevento e Caserta. Today, the fossiliferous
outcrop of Pietraroja is protected by a fence and is a
geo-palaeontological park.

Geological framework

The central-southern Italian Apennines are a thrust-
and-fold belt that originated during the Late Tertiary
from the deformation of the continental margin of the
Adria Plate. This plate is interpreted as being either an
independent Cretaceous unit or part of the Africa Plate
(Channel et a/., 1979). The Matese Mountains are part
of the Simbruini-Ernici-Matese structural unit (Patacca

& Scandone, 2007), that is made of 3-4 km thick Meso-
zoic carbonate platform deposits topped with Tertiary
carbonates and terrigenous sediments (D’Argenio et
al., 1973). During the Miocene, deformation piled the
Matese onto more external (eastern) domains; subse-
quent uplift occurred from the latest Pliocene to the
Pleistocene and brought the Matese to its present el-
evation (Carannante e? al., 2006).

In the Early Cretaceous, the present-day Pietraroja
area was part of a shallow-water carbonate domain
(Abruzzese-Campana Platform) which developed in
tropical-subtropical climatic conditions (D’Argenio,
1976). Although the regional palaeoceanography of
the continental margin to which Pietraroja belonged
is still debated, the stratigraphic context and evolution
of the Abruzzese-Campana Platform is well-assessed
(D’Argenio et al., 1973). As for other carbonate do-
mains of the Southern Tethyan Margin, predominantly
shallow lagoonal conditions prevailed during the Early
Cretaceous. Beginning with the Late Aptian-Early Al-
bian, margin retrogradation and broad tectonic uplifts
took place, with the establishment of more open ma-
rine conditions, including deeper margins connect-
ing the shelf to the basin areas (Vigorito ef al., 2003;
Carannante et al., 2004). According to the these au-
thors, between the Middle Aptian and Early Albian
the southeastern sectors of the Matese Mountains were
downfaulted, leading to the development of a narrow
eastward-facing channelised carbonate margin.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

ll

COMPLEX F
A mudstones and wackestones alternated to

marly and clayey, locally bituminous interlayers.

Ceratolithoides aculeus, Cretarhabdus
angustiforatus, Eiffellithus eximius,
Lucianorhabdus sp., Micula decussata,
Prediscosphaera cretacea,
Rucinolithus sp., Russellia sp..

E lo 0 EE, MITI
ST ComPLexE
fine laminated mudstones with thin strata
UPPER and lenses of packstones and wackestones
PLATTENKALK rich in spicules of siliceous sponges, which

prevails in the middle and upper part;

locally alternating with muddy layers;
rare macrofossils, no vertebrate remains.

radiolarians, Quinqueloculina sp.,
Pseudonummoloculina sp., Giomospira sp.,
Cuneolina spp., Sabaudia minuta,
Praechrysalidina infracretacea,
Mesorbitolina sp., Paracoskinolina
funesiana, and Cribellopsis arnaudae.

COMPLEX D

rudstones and floatstones rich in bioclasts,
intraclasts, peloids, and lithociasts, intercalated
with packstones and grainstones and
subordinately mudstones/Awackestones;
corals, ostreids, requienids, gastropods, and

- echinoids;
ox Cayeuxia, Ovalveolina reichelii,
La Salpingoporella dinarica, Praechrysalidina
® infracretacica, orbitolinids, Trip/oporella
Dv marsicana, Bacinella imegularis.
n°]
= COMPLEX C
a
(o)
2
2
Ko)
d SSATT@SSST= COMPLEX B
= T°2 an 7A
3 LO O A
“-Q 2

3 O Sd
ha Didi a Dn

lasciati nani

—— sn
c|150 ===
Rs
=
[ol
<T
di
z
(e)
E COMPLEX A
=
Rss)
= ==
(o) CSNTA
= aaa
(o) netta ce 2-4 7°
co {100 > — Piglia

rea: at-Id-2 2

{ NO DATA
COMPLEX A

Neocomian

continental plants, fishes, terrestrial vertebrates.

Burdigalian

Cc
DS)
Q
<q
da
®
>
(e)
dg

Tana
pi
a

|

il

detritic limestones

in thick strata;

briozoans, ostreids,

red algae (Lithothamnium).

miocenic transgressive
surface

INTERVAL D
PLATTENKALK Il
lithographic limestones
with thin levels and nodules
of chert;

Paracoskinolina tunesiana;
continental plants, fishes,
terrestrial vertebrates.

INTERVAL C

detritic and micritic
limestones in thin strata;
Cribellopsis cf. amaudae,
Paracoskinolina tunesiana.

INTERVAL B
PLATTENKALK I
lithographic limestones;
Cribellopsis cf. amaudae,
Paracoskinolina tunesiana,
Ovalveolina reichelii.

—E-

INTERVALA

detritic limestones

in thick strata;

Cribellopsis cf. armaudae,
Paracoskinolina tunesiana,
Ovalveolina reicheli.

Fig. $ - A) lithostratigraphic column of the Cretaceous succession of Pietraroja, according to Carannante et a/. (2006); B) stratigraphic
sequence of the north-eastern side of the “Civita di Pietraroja”, according to Bravi & Garassino (1998). Plattenkalk I and II in B
roughly correspond to lower and upper Plattenkalk in A. Abbreviations: U) unconformity; E) erosional surfaces. Microfossils in green,

macrofossils in brown. (Modified from the mentioned authors).

Fig. 5 - A) colonna litostratigrafica della successione cretacea di Pietraroja, secondo Carannante e? a/. (2006); B) sequenza stratigrafica
del versante nord-orientale della Civita di Pietraroja, secondo Bravi & Garassino (1998). I livelli denominati Plattenkalk I e Plattenkalk
II in B corrispondono approssimativamente ai livelli denominati lower Plattenkalk e upper Plattenkalk in A. Abbreviazioni: U) discor-
danza; E) superfici di erosione. Microfossili in verde, macrofossili in marrone. (Dagli autori citati, modificate).

12) CRISTIANO DAL SASSO & SIMONE MAGANUCO

Lithology, stratigraphy and sedimentology

The Cretaceous succession of Pietraroja, which is
about 340 m thick, includes two well-stratified Plat-
tenkalk horizons (Fig. 5). According to Carannante et al.
(2006), each Plattenkalk presents distinct fining and thin-
ning upward trends with a detritic basal layer. The lower
Plattenkalk is averagely coarser and is characterised by
the occurrence of frequent benthonic foraminifera-rich
grainstones and packstones: it is relatively poor in mac-
rofossils and, to date, no vertebrate remains have been
reported from this interval. Above this is a second Plat-
tenkalk horizon, with a depth of 8-9 m. The thickness of
the upper Plattenkalk increases to the southwest, reach-
ing a maximum of some 15 m at the Le Cavere outcrop,
which is the source of the major fossil finds. The upper
Plattenkalk is averagely finer when compared with the
lower one, as it is mostly composed of fine laminated
mudstones (Fig. 6). In the middle and upper part of this
interval there is a prevalence of mudstones with thin
strata and lenses of packstones and wackestones rich in
spicules of siliceous sponges. Locally, chert lenses and
nodules, as well as bituminous and coprolite-rich beds,
are also found. The trend of the layers is about N-S lead-
ing eastward, with an average dip around 20°. This trend

Fig. 6 - A section of the upper Plattenkalk series, outeropping at Le
Cavere a few metres north of the point where Scipionyx was likely col-
lected.

Fig. 6 - Una sezione della serie superiore del Plattenkalk, affiorante a Le
Cavere pochi metri a nord del punto in cui è stato raccolto Scipionyx.

is almost identical to that of the lower Plattenkalk, except
for a “flute-beak” termination, where it becomes more
slanted (Bravi & Garassino, 1998).

Carannante et al. (2006) recently dated the Pietraroja
Plattenkalk as Lower Albian on the basis of its microfossil
assemblage, which includes Bacinella irregularis, Glo-
mospira sp., Cuneolina aff. pavonia, Praechrysalidina
infracretacea, Nummuloculina sp., Thaumatoporella sp.,
Debarina sp., Ovalveolina reichelii, Sabaudia minuta,
orbitolinids, miliolids and textularids. This fits well with
previous biostratigraphic data provided by Bravi & De
Castro (1995) and Bravi & Garassino (1998), who re-
corded the following association: Quinqueloculina sp.,
Pseudonummoloculina sp., Glomospira sp., Cuneolina
sp., Sabaudia minuta, Praechrysalidina infracretacea,
Mesorbitolina sp., Paracoskinolina tunesiana and Cribel-
lopsis arnaudae. Furthermore, D°Argenio (1963) and
Catenacci & Manfredini (1963) reported planktonic fo-
raminifers, tentatively referred to as Praeglobotruncana
sp., from the upper portion of the Plattenkalk.

Analysis of selected samples from the Pietraroja Plat-
tenkalk conducted by Carannante e? al. (2006) has led to
the recognition of three main microfacies:

Biolithoclastic packstone/grainstone microfacies (1) -
this occurs as strata or lenses a few centimetres thick. It
makes up most of the lower Plattenkalk, but occurs also
in the upper Plattenkalk. Components are mainly carbon-
ate intra- and/or lithoclasts, benthic foraminifers as well
as green algae, rudists, corals and thin-shelled bivalves.
Microfossil assemblages likely include a mixture of fos-
sils of various ages, from Aptian to Lower Albian (Bravi
& Garassino, 1998), and show a conspicuous contribution
from the demolition of older deposits.

Spicule-rich packstone/wackestone microfacies (2) -
this occurs as 0.005-0.1 m thick lenses, laminae and/or
continuous strata dominating the lower-middle portions
of the upper Plattenkalk. Siliceous and rare carbonate
sponge spicules are the main components, together with
subordinate radiolarians. Rare planktonic foraminifers
have been also reported (Catenacci & Manfredini, 1963;
D’Argenio, 1963).

Thinly laminated mudstone microfacies (3) - this
makes up most of the upper portion of the upper Plat-
tenkalk. The deposits are an alternation of several-tens-
of-micrometers-thick light havana and ash-grey carbonate
muddy laminae. Locally, thin marly, cherty and rare dark
grey clay layers alternate with calcareous laminae. Mi-
crofossils are absent or rare and consist mainly of sponge
spicules.

Stratigraphic position of Scipionyx

According to the person who found the fossil (To-
desco, pers. comm., 1993), Scipionyx was collected at a
point of the Le Cavere quarry that, based on the present-
day topography, falls into the lower-right quarter of the
fenced-off area of the outcrop (Fig. 7). More precisely,
the discovery site is located to the right of the entrance
gate and some 7-9 m from the paved road (once a gravel
one). Todesco also has stated (pers. comm., 2010) that the
layer from which Scipionyx came from outeropped some
50-80 cm higher than the present stepping (bottom) lay-
er, but in any case at an elevation below the level of the

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 13

Fig. 7 - A) orthophotograph of Pietraroja, and the fenced-off area (black arrow) of Le Cavere north of the village; B) western view of
the Le Cavere quarry, showing the probable point where Scipionyx was collected (red arrow). Scale bar = 100 m.

Fig. 7 - A) ortofotografia di Pietraroja e dell’area cintata de Le Cavere, a nord del paese (freccia nera); B) vista da ovest della cava de
Le Cavere, con indicazione del probabile punto in cui è venuto alla luce Scipionyx (freccia rossa). Scala metrica = 100 m.

Fig. 8 - A) position of the fish remains co-occurring with Scipionyx samniticus; B-F) close-ups of five clupeomorph fishes (C/upavus
sp.); G) indeterminate vertebra; H) isolated Notagogus-like scale. Scale bar = 10 mm.

Fig. 8 - A) posizione dei resti di pesci associati allo scheletro di Scipionyx samniticus; B-F) particolari di cinque pesci clupeomorfi
(Clupavus sp.); G) vertebra indeterminata; H) squama isolata di un pesce simile a Notagogus. Scala metrica = 10 mm.

road. The same layer produced several fish remains, such
as pyenodontiform jaws and Diplomystus-like fins (To-
desco, pers. comm., 1993). Incidentally, the remains of 5
clupeomorph fish (C/upavus sp.) are embedded within the
slab containing Scipionyx. These are preserved to various
degrees, from fully articulated to scattered bone clusters
(Fig. 8). One isolated scale — referable to a larger species
of fish, possibly the semionotiform Notagogus — is locat-
ed 22 mm above the distalmost tip of the dinosaur’s tail
(Fig. 8). In our opinion, the position of this layer, as well
as the contents of the slab, are consistent with the lower-

middle series of the upper Plattenkalk (sensu Carannante
et al., 2006). Nevertheless, to confirm this, and to more
precisely correlate the fossil with the known stratigraphy
of Pietraroja, we analysed the lithological and sedimen-
tological aspects of the slab that contains Scipionyx. The
artifactitious puzzle created by the various pieces of slabs
originating from different layers, which were assembled
around and under the specimen (Figs. 9-10), and the elim-
ination of the overlying layers during the preparation of
the fossil, rendered the search for the authentic embed-
ding layer tricky.

14 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 9 - The artifactitious puzzle of slabs assembled around the speci-
men, highlighted under ultraviolet light by the bright fluorescence of
the glue used (lightest blue).

Fig. 9 - Il mosaico di lastre assemblate attorno all’esemplare, eviden-
ziato in luce ultravioletta dalla forte fluorescenza del collante impiegato
(azzurro più chiaro).

Fig. 10 - Distinction between the authentic bedding layer of Scipionyx
(outlined in black) and the slabs added by the collector (shaded in grey).
See also cover page. Samples of sediment (red dots) were carefully
taken from the original layer, along a section of the granular bed (ochre
yellow areas) once covering the dinosaur.

Fig. 10 - Distinzione tra lo strato originario in cui giace Scipionyx (delimi-
tato dalle linee nere) e le lastre aggiunte dal raccoglitore (velate di grigio).
Si veda anche la foto di copertina. I campioni di sedimento (punti rossi)
sono stati prelevati con attenzione dallo strato originario, lungo una sezione
del livello granulare (zone giallo ocra) che prima ricopriva il dinosauro.

Macroscopic aspect of the embedding sediment -
After mechanical preparation, the sediment of the bed im-
mediately underlying and partially containing the fossil
Scipionyx became the most exposed one. In other words,
the sediment seen surrounding the specimen does not rep-
resent the whole layer that originally contained the fossil,
but only a portion of it, as most of it was removed to bring
to light the fossil like a bas-relief. The exposed sediment
is mostly fine-grained, but shows some textural inhomo-
geneities in the form of localised accumulations of larger-
grained sediment. The latter is dominant in the overlaying
bed, which, as we infer from the fossil’s thickness, gave
a major contribution to encompassing it. Luckily, patches
of this sediment remain at some distance from the fos-

silised bones of the dinosaur: 40-50 mm above the dino-
saur’s pelvis; 15 mm above the skull; 10 mm below the
left forearm and the right knee. There, the larger grains
surface in three very oblique, almost longitudinal, sec-
tions of the bed, which on macroscopic examination ap-
pear as grey-dotted granular halos (Fig. 11). We examined
three samples of the section nearest to the fossil (Fig. 10)
with optical and scanning electron microscopy.

Cross sections (10-15 mm thick) of the original se-
quence of layers embedding and underlying Scipionyx are
visible along a couple of adjacent cracks that delimit the
sediment surrounding the forearms and the pectoral and
abdominal sides of the dinosaur. Dense, alternating light
and dark laminations thinner than 1 mm can be seen there
(Fig. 12). These laminations are similar to the light ha-
vana and ash-grey laminae that characterise the fish-rich
layers (Dal Sasso & Maganuco, pers. obs., 2003-2010 on
historical and recently collected specimens), and are con-
sistent with microfacies (2) and (3) described by Caran-
nante ef al. (2006).

Fig. 11 - Grazing view of the granular bed that was sampled and exam-
ined under scanning electron microscopy (red arrow). The white arrow
indicates the point of view of the cross-section shown in Fig. 12.

Fig. 11 - Vista radente del livello granulare campionato ed esaminato
al microscopio elettronico a scansione (freccia rossa). La freccia bianca
indica la posizione della sezione mostrata in Fig. 12.

Fig. 12 - Exposed cross-section of the original slab of Plattenkalk embed-
ding Scipionyx, and a supporting slab added by the collector (shaded in
grey). The red arrow points to the bed examined in Fig. 13; the white
arrow indicates the left manus of Scipionyx. Scale bar = 1 mm.

Fig. 12 - Sezione esposta della lastra originaria di Plattenkalk che racchiude
Scipionyx e di una lastra di supporto aggiunta dal raccoglitore (velata di
grigio). La freccia rossa indica il livello esaminato in Fig. 13; la freccia
bianca indica la mano sinistra di Scipionyx. Scala metrica = 1 mm.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 15

Microscopic aspect and chemical composition of
the embedding sediment - Under the light microscope,
the original sediment embedding Scipionyx has two main
components: a fine-grained (0.01-0.03 mm), white ma-
trix, probably composed of carbonates; and a medium-
grained (0.10-0.26 mm), vitreous, brown substrate, pos-
sibly siliceous clasts, that according to Torricelli (pers.
comm., 2010) are more or less fragmented sponge spi-
cules (Fig. 13B). Based on the classification of Dunham
(1962), this sediment can be defined as an intermediate
between packstone and wackestone in that it is matrix-
supported but consisting in more than 10% grains, some
of them in contact.

SEM element microanalysis ofthe three samples of sed-
iment confirmed that the fine-grained matrix is composed
of calcium carbonate, thus it can be defined as a micrite
(Fig. 13A-B); in contrast, the larger grains are composed
of silica, thus they are likely the fragmented spicules of
siliceous sponges (Fig. 13B-C). The low aluminium peak
indicates a secondary terrigenous component (clay miner-
als). Worthy of note is the total absence of phosphorus, one
of the main elements in the chemical composition of the
fossil. That means that the abundant phosphorus preserved
in Scipionyx does not come from the depositional environ-
ment but derives from the decay of the carcass. This is a
very important datum, which will be dealt with in more
detail in the section dealing with soft-tissue taphonomy.

Both the texture and chemical composition of our
samples of sediment fit well with the description of mi-
crofacies (2), the “spicule-rich packstone/wackestone
microfacies” (Carannante et al., 2006). Microfacies (2)
“largely dominate[s] the lower-middle portions of the
upper Plattenkalk”, providing good sedimentological
evidence that Scipionyx comes from one of those layers.
In our opinion, and according to Torricelli (pers. comm.,
2010), the probable presence of microfacies (3) in the bed
underlying Scipionyx, as well as in other stratigraphically
adjacent layers, strengthens, rather than contradicts, that

| Cursor=9 235 keV. 2cnt ID=
Vert=1000 Window 0.005 - 40955= 19252 ent

conclusion. Indeed, the thin strata and spicule-rich lenses
of packstones and wackestones typically intercalate with
mudstones in the lower and middle parts of the upper Plat-
tenkalk (Carannante et a/., 2006). At the same time, the
layers embedding Scipionyx cannot pertain to microfacies
(1) because none of our samples is grain supported (sensu
Dunham, 1962), and no foraminifers, algae, coral or bi-
valve fragments are present.

Other observations lead us to exclude another strati-
graphic portion of the Pietraroja outcrop: without doubt,
Scipionyx does not come from the uppermost series of the
upper Plattenkalk. In 2001, fieldwork was conducted by
the MSNM, in collaboration with the SBA-SA, to investi-
gate the fossil contents ofthe uppermost strata ofthe upper
Plattenkalk at the Le Cavere locality. Samples collected
Just a few meters from the point where Scipionyx was un-
covered, from the top of the series to a depth of 3 meters,
were numbered as a continuous series of 20 levels and
were analysed in parallel by two sedimentology laborato-
ries. Almost all these layers turned out to be dominated by
microfacies (3) as described by Carannante e? al. (2006):
very fine mudstone, with absent to rare microfossils. In
fact, palynology and calcareous nannoplankton analysis
revealed these layers to be completely sterile (Torricelli,
pers. comm., 2007; Maffioli, pers. comm., 2009).

Depositional environment

The exceptional degree of preservation of the fossil
vertebrates from Pietraroja results from peculiar deposi-
tional conditions. The relatively small size of the speci-
mens, the rapid sedimentation of fine-grained sediments
in a subaqueous environment, and the scarcity or absence
of oxygen, are among the physical parameters that al-
lowed the exceptionally fine fossilisation of bones and
soft tissues in this site (see also Soft Tissue Taphonomy).

Cursor=15.755 keY_ Ocnt ID=Polgl Zrkal Ub6
Vert=500 Window 0.005 - 40.955= 21826 cnt

Fig. 13 - Close-up of the sampled granular bed seen under the optical microscope, and SEM element microanalysis of its two main
components. A) elemental spectrum of the fine-grained matrix (white areas in B), composed of calcium carbonate; C) elemental spec-
trum of the medium-grained clasts (brown-coloured inclusions in B), composed of fragmented spicules of siliceous sponges. Note the
absence of phosphorus, which is one of the main components of the Scipionyx fossil. Scale bar = 0.1 mm.

Fig. 13 - Particolare al microscopio ottico di un campione del livello granulare e microanalisi degli elementi (al SEM) delle sue due
componenti principali. A) spettro degli elementi della matrice a grana fine (bianca in B), composta da carbonato di calcio; C) spettro
degli elementi dei clasti di grana media (marroni in B), che sono frammenti di spicole di spugne silicee. Si noti l’assenza di fosforo,
che invece è uno dei componenti principali del fossile di Scipionyx. Scala metrica = 0,1 mm.

16 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fossil-rich Plattenkalks have been documented from shal-
low lagoonal to relatively deep basinal settings, so there
is still great uncertainty on the depositional environment
that created the Pietraroja Plattenkalk. On this point, three
distinct models have been proposed.

Lagoon model - According to some authors
(D’Argenio, 1963; Bravi & Garassino, 1998), the Pie-
traroja Plattenkalk was laid down close to a coastal area,
in a very shallow lagoonal environment that was fre-
quently isolated from the open sea but subject to tidal
influence and occasional storms. “Apparent cyclicity in
the lithotype distribution” (Bravi & Garassino, 1998),
inferred from the graded horizons in most layers of the
upper Plattenkalk, led to hypothesise that sediment depo-
sition was linked to the rhythm of tides and storms which
reached the lagoons and poured the finer sediments of the
surrounding above-sea-level areas into them. Those in-
stantaneous microturbiditic events alternated with more
or less prolonged intervals of sedimentary starvation, that
Bravi & Garassino (1998) attribute to isolation of the ba-
sins from the open sea — these periods would have caused
overheating and dystrophic conditions in the waters of the
lagoons. The presence of bioturbations in the Pietraroja
limestone suggests that oxygen was not entirely lacking
on the seabed, whereas several layers particularly rich in
fish and coprolites document mass mortality events. Ac-
cording to this model, films of algae and bacteria would
have been the most probable origin of the fine laminations
seen in these layers.

According to D’Argenio (1963), putative mud cracks,
cracked chips and gas pits in some Plattenkalk layers indi-
cate occasional emersions. However, Bravi & Garassino
(1998) did not find any certain sign of the drying of sedi-
ments in the same stratigraphic series. On the other hand,
the latter authors regard the abundance of sponge spicules
in several layers as not necessarily linked to a deep depo-
sitional environment, since sponge colonies can develop
also in very shallow water. According to Bravi & Gar-
assino (1998), strong overheating of the waters of the Pie-
traroja lagoons, probably occurring when they came close
to the limit of emersion, is proven by certain layers of the
upper Plattenkalk that are literally carpeted with bivalves,
all lying with the valves open upwards and still articulated
at the umbones. The lack of evaporitic minerals, expected
to be present in overheated coastal lagoons, is explained
by occasional freshwater contribution. In this perspective,
the exceptional integrity of the vertebrate fossils is linked
to calm lagoonal waters, and the random bone dispersion
of disarticulated skeletons is linked to bioturbation rath-
er than to currents. Conversely, the mutilation of some
carcasses and the dorsally arcuate position of many fish,
caused by putrefaction gases in the abdominal cavity, are
linked to flotation. From this, it is inferred that the depth of
the basins was limited, because putrefaction gases might
not expand the abdomen of dead fish at higher pressures
(Bravi & Garassino, 1998). In any case, Scipionyx suf-
fered a different fate: floating carcasses never reach the
bottom intact, so we can definitely conclude that the little
dinosaur was drawn down to the bottom rapidly, and just
as rapidly buried.

Slope/shallow basin model - Contrary to the authors
above, Catenacci & Manfredini (1963) believe that depo-

sition of the Pietraroja Plattenkalk took place in a coastal
strip which ran between the margin of the platform and
the deepest sediments of the opposite basin (Molise-San-
nitica depression). The transitional nature of this environ-
ment is inferred by sedimentological and stratigraphic
observations, including a lateral ‘flute-beak” transition
from reefoidal to plattelkalk deposits. This type of depo-
sitional environment would have been subjected to recur-
rent bathymetric variations causing the irregular alterna-
tion of semicontinental-lagoonal conditions and open-sea
conditions. The authors explain in this way the presence
of very different organisms and lithotypes (amphibians,
reptiles and crustaceans on one side; radiolarians, sponge
spicules, fishes and chert on the other). Such “alternation”’
of very different domains seems, however, very unlikely.
Freels (1975) hypothesised that the Pietraroja Plat-
tenkalk was deposited in a submarine erosion basin char-
acterised by low-energy and reducing conditions towards
the sea bottom (stagnation-type deposits sensu Seilacher,
1970). From the geometry of the Plattenkalk bodies, and
from the way in which they overlap the adjacent carbon-
ate platform facies, Freels (1975) estimated the Pietraroja
basin to have been about 60 m deep and 1 km wide. The
filling of the basin would have been due to suspension
currents developing at the margin of the basin and carry-
ing materials derived from the adjacent depositional envi-
ronments. Evidence of such suspension currents would be
provided by slumpings at the borders of the Plattenkalk
and by gradation of some layers of the Plattenkalk it-
self. Consistent with this hypothesis, the structures that
D’Argenio (1963) interpreted as caused by sediment dry-
ing would be ascribed, rather, to subaqueous shrinkage.

Submarine channel model - Both models above ap-
peared inadequate in that, according to some researchers,
they failed to take into account all the sedimentological
and palaeontological features ofthe Pietraroja Plattenkalk.
A new model was therefore proposed more recently by
Carannante ef al. (2006). According to this, the Pietraroja
Plattenkalk sequences are not shallow lagoonal depos-
its or intra-platform basin-fill, but deposits of a subma-
rine channel (referred to as the Pietraroja Channel) that
document a major transgressive event. In particular, the
lower Plattenkalk and the fine-grained, fossil-rich upper
Plattenkalk are interpreted as representing, respectively,
channel-fill and abandon deposits, which followed previ-
ous coarse-grained channel-fill sequences as a response to
the demise ofthe channel as a sedimentary conduit. Caran-
nante ef al. (2006) support this model with a number of
observations. First of all, these authors note that the Pie-
traroja Plattenkalk is arranged within a relatively narrow,
east-dipping incision, and is confined laterally by coarse
biointraclastic deposits and by a sharp erosive surface.
Secondly, both the lower and the upper Plattenkalk con-
sist in moderately to well-sorted deposits. It is improbable
that these could form in upper subtidal to intertidal la-
goonal settings like those proposed by D’Argenio (1963)
and Bravi & Garassino (1998); rather, they usually origi-
nate from carbonate turbidites. Palaeocurrent measure-
ments have been reported to be documented in the two
Plattenkalks by scours, groove marks, oriented grains and
ripples, and to indicate sediment transportation always to-
wards the east. However, Carannante e? a/. (2006) did not
find any evidence for multi-directional or bi-directional

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY [7

flows or for turbidite ponding, that characterise wave-
dominated and tide-dominated shallow-water environ-
ments and small intraplatform basins. Therefore, it is
likely that microfacies (1) derives from eroded submarine
sediments that were re-arranged mainly as turbidites or
bottom current-related deposits; microfacies (2) derives
from primary deposition from bottom currents and tur-
bidites; and microfacies (3) originates from deposition of
low-density, muddy turbidites. According to this model,
we deduce that Scipionyx was buried in a single, rapid
event by a turbidite.

The fossils of Pietraroja range from the beautifully
preserved Scipionyx to very poorly preserved and uniden-
tifiable skeletal remains. According to Carannante et al.
(2006), these features can be recognized both at outcrop
scale and over a small area of a single bed, and may mean
that the carcasses came from different places and were
subjected to different transportation processes. Moreover,
the fossils seem to increase in dimension moving from
the borders of the inferred Pietraroja Channel towards the
axis. This spatial arrangement may derive from the kind
of transportation that was able to drag or carry large car-
casses only over limited distances.

Carannante e? al. (2006) also observe that the inverte-
brate fauna assemblage is very poor compared to the ver-
tebrate one. It is unlikely that this could have been caused
by a taphonomic artefact, because the exoskeleton of most
of the site’s invertebrate taxa would have rendered them
more resistant than the vertebrates to decay. As a matter of
fact, Bravi & Garassino (1998), while supporting the la-
goon model, describe the decapod crustacean assemblage
of Pietraroja as having “a not particularly good state of
preservation”: in fact, only 17 of the 29 studied specimens
could be given a confident systematic ascription. A_ very
similar conclusion was reached a century ago on different
decapod specimens by D’Erasmo (1914, 1915).

According to Carannante e? al. (2006), and contra oth-
er authors (D’Argenio, 1963; Bravi & Garassino, 1998),
the bioturbation index of the site is very low, most traces
being ascribed to repichnia, with no evidence of domich-
nia or pascichnia. This observation is consistent with a
shallow, anoxic surface very close to the water-sediment
interface. Anoxic to suboxic conditions at the seafloor fit
well with the occurrence of calcareous or marly bitumi-
nous layers. They also explain the very scarce benthic
invertebrates, and a curious lack of nematodes, annelids
or any other “worm”, which in any marine biocoenoses
(including lagoons) are quite abundant. Also, according to
Carannante et a/. (2006) the finding of very small bivalves
preserved with open but still connected valves cannot be
interpreted as a mass-death event caused by overheating
water (Bravi & Garassino, 1998); rather, this is another
typical feature of anoxia, given that only under this condi-
tion does the hinge ligament not decompose — thus keep-
ing the two valves conjoined — and the adductor muscles
relax — opening the bivalves up.

Another interesting point is the strong presence of du-
rophagous fish coupled with the almost complete absence
of possible prey. If the site represents a thanatocoensis, as
previously thought, then a large number of shelled inver-
tebrates, or at least of corals, should be present. On the
contrary, a large part of the fish fauna is composed from
pelagic predators, like Belonostomus and at least two spe-
cies of ichthyodectid fishes (Signore ef a/., 2005). So, the

faunal assemblage of Pietraroja likely represents a mixture
of organisms originating from different types of environ-
ment, but certainly not from a lagoonal one. Summing
up, Carannante e? a/. (2006) concluded that the fossils of
the Pietraroja Plattenkalk represent a taphocoenosis and
an obruption deposit (sensu Bottjer et a/., 2002; Martin,
1999). This scenario is also supported by the scarcity of
terrestrial animals and the complete absence of insects,
which are found in fossil lagoons from the Carboniferous
onwards. Since there is no sedimentological evidence of
a fluvial contribution, it is conceivable that the terrestrial
taxa were drawn into the basin during exceptional events,
such as storms and hurricanes. Whether Scipionyx drowned
or died from other causes, remains impossible to say.

In our opinion, further work (especially in the field) is
needed to evaluate the models above. For example, some
crucial information supporting the model of Carannante
et al. (2006), such as the lack of ichnofossils and the size-
related distribution of body fossils on the surface of the
strata, are based on observations made on a quite limited
sampling area, which was “too small to offer conclusive
evidence” (Signore, 2004). In addition, as admitted by Si-
gnore (2004), who co-authored the paper of Carannante ef?
al. (2006), this sampling area was excavated rapidly be-
cause of an urgency to terminate the building of the base-
ment of a water reservoir.

In the taphonomic study of Scipionyx we actually no-
ticed some difference from the Solnhofen fossils (see Soft
Tissue Taphonomy). In particular, the well-documented
evidence that phosphorus is limited to the body of Scipio-
nyx (1.e., not present in the sediment) has lead us to prefer
a depositional environment of relatively shallow, open
marine waters, rather than a lagoon where phosphates de-
riving from organic activity would have more easily ac-
cumulated. Although in a lagoon some additional factors
— such as reduced vertical water remix and seasonal over-
heating — might favour low oxygen levels on the sea bed
(Bravi & Garassino, 1998), anoxia is not at all limited,
nor linked, to “closed” and/or shallow basins. Rather, it
is worth to mention that in aquatic settings under normal
conditions (1.e., open waters), the oxic-anoxic boundary
may be at, or even above, the sediment-water interface,
and that under water, most decay of carcasses is usually
anaerobic (Briggs, 2003). For sure, the complete and rapid
burial of Scipionyx in a relatively soft, muddy substrate at
a single depositional event would have been conducive in
greatly increasing the chances of soft tissue preservation,
in saving the carcass from violent crushing, and in permit-
ting subsequent slow, plastic, diagenetic compaction.

The fossil assemblage and palaeoenvironment

Since the site was first reported more than two centu-
ries ago (Breislak, 1798), hundreds of fossils have been
recovered from Pietraroja. The biodiversity of the site, re-
vealed by the number of taxa found there (e.g., Fig. 14),
indicates that the depositional basin was surrounded by a
variety of natural settings. Below, we give a concise list of
the flora and fauna so far described.

Microfossils - The site contains a significant quantity
of radiolarians and spicules from siliceal sponges. Their
abundance is documented also indirectly by the chert lay-

18 CRISTIANO DAL SASSO & SIMONE MAGANUCO

ers and nodules derived from the accumulation and natu-
ral processing of the skeletal remains of these organisms.
Benthic macroforaminifers are also found in the lower and
upper Pietraroja Plattenkalks. These include orbitolinids,
alveolinids, miliolids and agglutinated foraminifers (Cat-
enacci & Manfredini, 1963; D’Argenio, 1963; Bravi &
De Castro, 1995; Bravi & Garassino, 1998; Carannante e?
al., 2006; for a list of all taxa, see Lithology, Stratigraphy
And Sedimentology). Planctonic foraminifers (Praeglo-
botruncana sp.) are rarer. Some layers are rich in calcare-
ous algae (Dasycladaceae).

Plants - Fossil plants are very rare, but they are very
useful palacoenvironmental indicators. The lower portion
of the upper Plattenkalk includes laminated horizons that
are richer in bitumen than the regular layers and that often
contain remains of terrestrial plants (Bravi & Garassino,
1998). At least two kinds of gymnosperms are represented
at Pietraroja: the Cheirolepidiaceae and the Bennettitales.
The former consist of loose branches belonging to the
araucaria-like conifer Brachyphyllum; the Bennettitales
are represented by abundant disarticulate foliage belong-
ing to the genus Podozamites, which shows some resem-
blance with extant cycads. The first fossil fern, referred
to the genus Ph/ebopteris (Matoniaceae), was recently
reported (Bartiromo et al., 2006).

Based on the different palaeobiology of these taxa, we
can hypothesise the gross distribution of the vegetation
with respect to the depositional basin. Probably, bennet-
titaleans and ferns flourished in proximity to the seashore,
whereas cheirolepidiaceans, capable of inhabiting more
arid settings, grew further inland. The leathery leaves of
the cheirolepidiaceans, and the other strategies against
evotranspiration exhibited by these plants, suggest that
some 110 million years ago the Pietraroja region had a
warm, dry climate for most of the year.

Invertebrates - The invertebrate records are abundant
and include a dozen mollusc species, such as Nerineidae
gastropods (Bravi, 1994; Marramà, pers. comm., 2007),
undetermined bivalves (Bravi, 1999) and Hoplitaceae
ammonites of the genus 7roch/eiceras. The presence of
these taxa confirms the age of the Pietraroja Plattenkalk
(Barbera & La Magna, 1999). Echinoderms are rare (Fig.
14A) and represented only by small-sized Asteroidea and
Ophiuroidea (Bravi, 1999; Freels, 1975). Interestingly, 3
brand new taxa of decapod crustaceans were described
by Bravi & Garassino (1998): the penaeid Micropenaeus
tenuirostris, the caridean Parvocaris samnitica (Fig. 14B)
and the thalassinid Hux/eycaris beneventana. A possible
palaemonid was mentioned by Signore (2004). Other fos-
sil specimens, which have not yet been examined in de-

Fig. 14 - A snapshot of the fossil biodiversity at Pietraroja: A) an indeterminate sea star; B) a decapod crustacean (Parvocaris sam-
nitica), C) a pyenodontid fish (Oc/oedus costai); D) a guitarfish (Ri inobatus obtusatus). Not to scale.

Fig. 14 - Esempi di biodiversità tra i fossili di Pietraroja: A) una stella marina indeterminata: B) un crostaceo decapode (Parvocaris
samnitica), C) un pesce picnodonte (Oc/oedus costai); D) un pesce chitarra (RAinobatus obtusatus). Proporzioni non rispettate.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 19

tail, indicate the presence of astacid crayfish (Bravi, 1999;
Marramà, pers. comm., 2007).

Fishes - Fishes are the most abundant and diversified
fossil vertebrates at Pietraroja: at least 20 species belong-
ing to 12 suprageneric taxa of cartilaginous and bony fish-
es have been recognised (for a complete list, see Bravi,
1999). Most coprolites, which abound in some layers of
the upper Plattenkalk, can be referred to fish.

The Osteichthyes are represented by basal neoptery-
gians (Holostei) and more derived neopterygians (Tele-
ostei). The former include three extinct suprageneric taxa:
the Pycnodontiformes (Fig. 14C), to which the genera
Paleobalistum (D’Erasmo, 1914, 1915) and Oc/oedus
(Poyato-Ariza & Wenz, 2002) belong, have a disk-shaped,
laterally flattened body, well-adapted for swimming in
narrow reefs, and typical dome-shaped crushing teeth,
which allowed them to feed on hard-shelled molluses and
corals. The Macrosemiiformes were small bony fish that
were very common in the Cretaceous oceans of the world;
at Pietraroja they are represented by Notagogus pentlandi,
a species with a peculiar dorsal fin composed from two
closely spaced lobes (Bravi, 1994). The Semionotiformes
include the genus Lepidotes, the members of which were
covered by very robust, diamond-shaped ganoid scales
and are regarded as freshwater forms.

At least six suprageneric teleostean taxa are represent-
ed at Pietraroja. The Aspidorhynchiformes, such as Be-
lonostomus crassirostris (Costa, 1853-1864), were slen-
der-bodied pelagic predators — similar to extant needlefish
— equipped with a long rostrum and pointed teeth. The
Ichthyodectiformes, open-water predators of even larger
size, are represented by the genus Chirocentrites and by
a second ichthyodectid (Signore ef al., 2005) recently
referred to the genus C/adocyclus (Signore et al., 2006).
Other predatory teleosteans include the elopomorph Ana-
ethalion and the possibly endemic pholidophoriform
Pleuropholis decastroi (Bravi, 1988), which is regarded a
freshwater-brackish form. Another species known only in
this locality is the ionoscopiform /onoscopus petraroiae
(Costa, 1853-1864). The Clupeomorpha are represented
by two genera: Diplomystus, an unspecialised form with
a very convex abdomen - related to present-day sardines
— and Clupavus, which is one of the smallest and most
abundant fishes in the Pietraroja outerop. The Chondrich-
thyes are represented by a fossil guitarfish, RAinobatus
obtusatus (Costa, 1853-1864), of which the only speci-
men so far recovered is remarkably articulated and well-
preserved, skin remains included (Fig. 14D). This is a rare
case, as the cartilaginous skeletons of the Chondrichthyes
have much less chance of becoming fossilised than their
bony teeth and scales.

As stated above, the submarine channel model would
more easily explain the downstream occurrence of spe-
cies belonging to quite different settings located at higher
elevations (open sea/reef/lagoon/brackish/freshwater). In
the lagoon hypothesis, on the other hand, one could as-
sume that fish inhabiting open, deep waters reached the
lagoon sporadically during catastrophic events.

Tetrapods - Although the upper Plattenkalk of Pietraro-
ja is historically renowned for its well-preserved ichthyo-
fauna, some tetrapods, including amphibians and reptiles,

had also been found before the discovery of Scipionyx.
The amphibians are represented by a single specimen, the
Batrachian Ce/tedens megacephalus (McGowan, 2002),
belonging to the Albanerpetontidae — an extinct group
distantly related to extant salamanders. Leathery skin,
well-ossified skeletal elements, and solidly built girdles
and limbs, indicate that albanerpetontids were primarily
land-dwelling animals (McGowan & Evans, 1995).

Lepidosauromorph and archosauromorph reptiles are
represented, respectively, by sphenodontian and squa-
mate lizards, and by still undescribed crocodylomorphs.
The Sphenodontia of Pietraroja include two different taxa,
represented by two single specimens: the first one (Fig.
15) is Derasmosaurus pietrarojae (Barbera & Macuglia,
1988); the second one is still unnamed, but in turn pre-
serves the remains of a small lizard, which was identified
and named EZichstaettisaurus gouldi, within its abdominal
cavity (Evans et al., 2004). Besides Eichstaettisaurus, two
other Squamata are known from the Pietraroja outcrop:
Chometokadmon fitzingeri (Costa, 1853-1864), similar in
body shape to modern scincids (Fig. 16), was considered
for many years a sphenodontian, but is in fact related to
the Anguimorpha, as established by a combination of cra-
nial and posteranial characters and recent cladistic analy-
sis (Evans et al., 2006); and Costasaurus rusconii (Estes,
1983), represented again by a single specimen, which was
a small-bodied, short-limbed animal thought originally to
be an amphibian. A fourth lizard taxon awaits description.

The Crocodylomorpha of Pietraroja consist in at least
two semi-complete, articulated specimens and in scattered
vertebrae that probably pertain to a single Mesoeucroco-
dylia taxon. A peculiar feature shared by all these remains
is their relatively small size, suggesting that these animals
did not exceed 1.5 meters in length (Dal Sasso, pers. obs.,
2001) — a similar size to extant dwarf crocodiles, dwarf
caimans and the Chinese alligator.

The skull fragment that was recently tentatively re-
ferred to a juvenile pterosaur (Signore, 2004: fig. 18) must
be regarded as such with caution in that it is very similar
in shape and size to the lower jaw of Bel/onostomus, an
aspidorhynchiform fish collected during the 2001 MSNM
fieldwork and very common at Le Cavere (Dal Sasso,
pers. obs., 2007; Marramà, pers. comm., 2007).

Fig. 15 - The holotype of Derasmosaurus pietraroiae (MPN 541), the
best preserved sphenodontian from Pietraroja. Scale bar = 10 mm.

Fig. 15 - L’olotipo di Derasmosaurus pietraroiae (MPN 541), lo sfeno-
dontide meglio conservato di Pietraroja. Scala metrica = 10 mm.

20

Lar 5°

Saly. Calò. dis.

VOI

CRISTIANO DAL SASSO & SIMONE MAGANUCO

Tav }l

Baf fadente ine.

Fig. 16 - Lithography illustrating the holotype of the squamate Chometokadmon fitzingeri (MPN 539) and its integumentary remains,

well-preserved on its skull and neck. (After Costa, 1853-1864).

Fig. 16 - Litografia che illustra l’olotipo dello squamato Chometokadmon fitzingeri (MPN 539) e i suoi resti di pelle, ben conservati

sul cranio e sul collo. (Da Costa, 1853-1864).

Given that marine batrachians have never been report-
ed, Celtedens is a truly important palaeoenvironmental in-
dicator. The presence of terrestrial reptiles, along with ter-
restrial plants and fish genera found also in ancient fresh-
water settings (e.g., Lepidotes, Pleuropholis), confirms
that dry land with surface freshwater was not far from the
depositional basin of Pietraroja. This land must have been
quite vast and persistent over time, because only under
these conditions would it have been possible to create the
ecological balances and complex food chains which had
predatory dinosaurs at the top.

Palaeogeography

Palaeogeographical data related to the Albian age sup-
port the hypothesis that Scipionyx and the other terrestrial
fauna of Pietraroja inhabited temporarily isolated lands
that probably rose up in the Cretaceous Tethys during the
Middle-Upper Aptian tectonic phases (Carannante et al.,
2006). Those lands were part of the Adria Plate and of the
Periadriatic Domain, the continental margin of the cen-
tral-western Tethys which included the present-day Adri-
atic Sea, Italian peninsula, Sicily, northern Tunisia, Malta,

Albania, Montenegro, Bosnia and Herzegovina, Croatia,
Slovenia and the Periadriatic orogenic belt (Channel et
al., 1979; Zappaterra, 1990).

Cillari et al. (2009) have outlined the two main mod-
els for the Mesozoic geodynamic evolution of the Adria,
which has been considered either an independent micro-
plate or an African Promontory. In the first model, the Io-
nian Tethys is connected with the Alpine Tethys separat-
ing the Periadriatic region from Africa during Jurassic and
Cretaceous times (Finetti, 2005). In contrast, the second
hypothesis envisages two independent oceanic systems
(Alpine Tethys and Ionian Tethys) divided by a continen-
tal crustal sector (Channel et a/., 1979; Rosenbaum et al.,
2004; Stampfli, 2005). In any case, the Periadriatic Do-
main was a complex puzzle of small units traditionally
described as an archipelago of carbonate platforms that
was well-separated from both Gondwana and Laurasia
(Dercourt et a/., 1993, 2000; Yilmaz et al., 1996; Patacca
& Scandone, 2004, 2007; Zappaterra, 1990, 1994). These
platforms (i.e., the Apenninic Platform, Apulian Plat-
form, Panormide Platform, Adriatic-Dinaric Platform,
Kruja Platform, Mirdita Platform and Gavrovo-Tripolitsa
Platform) have been interpreted as being similar to the
present-day Bahama Banks, and described as small (from

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 21

a few hundred to thousands of square kilometres), elon-
gate, emergent islands bordered by large, shallow lagoons
and biogenic margins and ramps, and separated by deep
pelagic basins (1.e., the Lagonegro-Molise, Sicano, Um-
bria-Marche and Est Gargano basins, and the Genzana-
Greco and Pindos troughs).

Within this scenario, the carbonate platform bearing
the present-day Pietraroja — the Apenninic (or Laziale-
Abruzzese-Campana) Platform — is assumed to have
been separated in Aptian-Albian times from the adjacent
Apulian Platform by the Lagonegro-Molise Basin. Evi-
dence of this basin — in the form of a narrow, deep seaway
that would have been able to produce isolation — is pro-
vided by turbidite and debris-flows deposits containing
the macroforaminiferan Orbitolina texana, and therefore
related to the Late Aptian-Early Albian age (Cati et al.,
1989; Zappaterra, 1990, 1994). The Apenninic Platform,
not exceeding 250x200 Km in size (Nicosia et al., 2007),
can thus be imagined as a small island no larger than the
present Sardinia. However, assuming the portion bearing
the present-day Pietraroja as having been a distinct subu-
nit (Abruzzese-Campana Platform; see D’Argenio er al.,
1973; Carannante er a/., 2006), the island could have been
even smaller, like Corsica (Dal Sasso 2001, 2004).

In this geographical and environmental framework,
the existence of compsognathid theropods and similar
small-sized terrestrial vertebrates is acceptable, but the
presence of animals such as sauropods and medium-large
theropods (Nicosia et al., 2007; see below) would seem
to represent a problem. Actually, some sauropods, such
as the South American titanosaur Neuquensaurus austra-
lis (body length 7-9 m, estimated body mass 3,500 kg)
and the German basal macronarian Europasaurus holgeri
(estimated adult body length 6.2 m, body mass 800 kg),
reached relatively small adult body sizes (Stein et al.,
2010). In particular, the Upper Cretaceous (Maastrichtian)
continental formations of the Hateg Basin of Romania
contain an array of relatively small-bodied dinosaur taxa,
including the titanosaurian sauropod Magyarosaurus da-
cus (body mass less than 1,000 kg), the basal hadrosaurid
Telmatosaurus and two species of the noniguanodontian
euornithopod Za/moxes. These, and other taxa such as the
Santonian hadrosauroid 7ethyshadros from Italy (Dalla
Vecchia, 2009) and the Hungarian ceratopsian Ajkacera-
tops (Osì et al., 2010), illustrate that dinosaurs, just like
Mediterranean dwarf proboscideans, were not exempt
from general ecological principles limiting body size
(Stein et al., 2010; and references therein).

Sedimentological evidence for long-lasting exposure
of dry land in the Matese area is provided by a strati-
graphic gap, which at the maximum amplitude spans from
the Albian p.p. up to the Turonian p.p. times, and by vast
deposits of bauxite, which also are Lower Albian in age
(Bravi & Garassino, 1998; Carannante et a/., 2006). Actu-
ally, the deposition of the fossiliferous upper Plattenkalk
deposits of Pietraroja preceded the Albian regressive
event and the formation of the bauxite bedrocks (Caran-
nante et al., 1988).

Evidence from central and southern Italian dino-
saur tracksites - Parallel studies of dinosaur tracksites
in central and southern Italy — all discovered in recent
years (Andreassi et al., 1999; Gianolla et al., 2000a,
2000b, 2001; Nicosia ef a/., 2000a, 2000b, 2007; Conti

et al., 2005; Sacchi et al., 2006, 2009; Petti ef a/., 20082,
2008b; Cillari ef a/., 2009) — are providing information
that may contribute towards a better understanding of the
palaeogeography of the region inhabited by Scipionyx.
These tracksites document the occurrence of a well-devel-
oped dinosaur fauna not only in the Apenninic (Laziale-
Abruzzese-Campana) Platform, but also in some adjacent
carbonate platforms of the Periadriatic Domain (Apulian
Platform, Panormide Platform), at least from Tithonian to
Santonian times.

As outlined above, most palaeogeographical recon-
structions represent the Adria Plate and the Periadriatic
Platforms as isolated areas separated from main continen-
tal dry lands by deep pelagic basins from the Pliensbachi-
an to the Coniacian — a long interval of about 130 My
(Nicosia ef al., 2007). Nevertheless, the new palaeonto-
logical finds in central and southern Italy indicate that the
Periadriatic Platforms were never completely separated
by deep seaways (Nicosia et al., 2007). Co-occurrence
of analogous dinosaur assemblages in different platforms
led also to hypothesise that some sort of geographical
connection above sea level existed, at least for certain pe-
riods. For example, the Latial ichnosite of Esperia (Petti
et al., 2008b) is coeval with the Apulian dinosaur track-
site of Bisceglie (Sacchi et a/., 2006), and both show the
presence in Aptian times of theropods and sauropods with
similar characters. Such faunal affinity was recently inter-
preted as an evidence of there having been at least tempo-
rary land bridges between the Apenninic and the Apulian
platforms during the Aptian (Petti ef a/., 2008b). Under
this perspective, the Laziale-Abruzzese part of the Apen-
ninic Platform is considered as having been a promontory
of the Apulian main bank, which probably acted as a bar-
rier between the northern Umbria-Marche Basin and the
southern Lagonegro-Molise Basin. This suggests that the
ancestors of Scipionyx, and maybe Scipionyx itself, would
not have been isolated from the neighbouring platforms
for a long time.

Similarly to Sacchi et al. (2009), we point out that the
present knowledge of dinosaur occurrence in the Italian
Periadriatic platforms is surprising and in definite contrast
with the idea of a continuous marine environment. The
amount of evidence on subaerial emergences — already
well-highlighted by a long list of data in a recent paper
(Nicosia et al., 2007) — increases considerably when the
whole Periadriatic region, including the Istrian Peninsula
and Croatia, is considered (Mezga & Bajraktarevié, 1999;
Dalla Vecchia, 2002, 2009; Mezga et al., 2006, 2007).
Thus, a model is needed that takes into consideration the
timing of the arrival of dinosaurs in central and southern
Italy, their possible immigration route and the possibility
for dinosaur assemblages to persist in confined areas. This
was attempted by Sacchi et a/. (2009), using the footprint-
bearing Aptian levels at Bisceglie (Apulia, southern Italy),
which occur intermittently within exclusively marine suc-
cessions. The presence of the dinosaur footprints could
be explained by two completely different hypotheses:
they might represent either the remaining traces — biased
by preservation windows — of a long-persisting autoch-
thonous association, or evidence for repeated immigration
events that were influenced by a filtering bridge. Anal-
ysis of all the available evidence, such as the scattered
occurrence of the tracks, the diversity of the dinosaurs,
the type of recorded environment, the type of diets, the

DI CRISTIANO DAL SASSO & SIMONE MAGANUCO

absolute and relative dimensions of the track makers and
the dimension of the available areas (Sacchi et a/., 2009),
tends to exclude the presence of a long-lasting coevolved
association. It does support, however, the occasional co-
occurrence of taxa taking place on account of separate mi-
gration events occurring over a period of least 70 million
years: in this model, dinosaurs would have been able to
immigrate into the carbonate platforms of the Periadriatic
area but not to survive there for a time long enough to al-
low for their co-evolution.

Co-evolution seems to be questioned also by a second
example. The Apulian dinosaur ichnocoenoses of Matti-
nata (Late Jurassic), Borgo Celano (Early Cretaceous) and
Altamura (Late Cretaceous) differ from one another but
correspond in faunal composition and evolutionary level
to age-equivalent dinosaur communities known from Eu-
rope, central Asia and North America (Conti et a/., 2005).
Therefore, it is difficult to accept the hypothesis of a par-
allel evolution on the mainland and in isolated areas.

In this scenario, the temporary land connections that
permitted the dispersal of dinosaurs within the area are
considered by Sacchi et a/. (2009) as real filtering-bridg-
es. The repeated occurrence of the complex biota known
to have occurred supports the hypothesis that there were
large, directly linked, unstressed biological “reservoirs”
that could re-inject plant and animal populations after
local extinction events. These ephemeral filtering-bridg-
es, while leaving seaways for the east-west spreading
of marine animals, would have allowed ‘“pre-adapted”
dinosaurs to reach the Periadriatic platforms in a north-
south route. Moreover, repeated connections among the
platforms themselves are hypothesised to have possibly
acted as a physical and environmental continuum, i.e. as
walkways between a land mass and the Periadriatic car-
bonate platforms during most of the Cretaceous (Sacchi
et al., 2009). The same dispersal mechanism is suggested
also by Dalla Vecchia (2002, 2009), albeit with another
terminology, when he interpreted these ephemeral islands
as having been stepping stones for “island hopping” of
dinosaurs across the Tethys.

Taking the filtering-bridges model as the most prob-
able, the problem of the dinosaurs’ origin remains open
— where did Scipionyx, and other Italian dinosaurs of
the Cretaceous, come from? The fact that the Periadri-
atic platforms were certainly bordered to the north by the
Ligure-Piemontese Ocean at least until the Turonian (e.g.,
Mindszenty et al., 1995), places the “homeland” of Italian
Cretaceous dinosaurs along the northern margin of Gond-
wana. For the same reason, putative connections with
Laurasia (Evans et al., 2004) or alternative north-south
routes via Iberia (Sereno ef al., 1994) are very unlikely
(Nicosia et al/., 2007; Canudo et al., 2009). In fact, cur-
rent understanding dates the last connection of the Adria
Plate to Laurasia — before their ultimate aggregation in the
Late Cretaceous — at Jurassic times, during the Bathonian
(Schettino & Scotese, 2002).

These constraints gave rise to new palaeogeographical
models of the western Tethys (Bosellini, 2002; Dalla Vec-
chia, 2002, 2005; Conti et al., 2005; Nicosia ef al., 2007;
Turco et al., 2007; Petti et a/., 2008; Sacchi ef al., 2009),
most of which incorporate a possible filtering bridge be-
tween the African continent and the Periadriatic platforms
(Fig. 17). Such a hypothesis has been recently refreshed
after the description of a titanosaur-like dinosaur track in

the Cenomanian of Sezze (Latina, central Italy), which
is part of the Apenninic Platform (Nicosia et al., 2007),
and the very recent discovery of a theropod bone in Ceno-
manian peritidal levels of Capaci (Palermo, Sicily), an
area that belongs geologically to the Panormide Platform
(Garilli et al., 2009). The geodynamic evolution of the
Panormide Platform, as inferred by Zarcone & Di Stefano
(2008), as well as the dinosaur fossil record of the region
(Petti er al., 2008b; Cillari et a/., 2009), is consistent with
a crustal sector of land connecting Africa to Adria above
sea level, and separating the Ionian Tethys from the Al-
pine Tethys during most of the Cretaceous period.

Palaeobiogeographical remarks - The palaeogeo-
graphical scenario of a variable number of islands ar-
ranged in a “string of pearls” between the Gondwanan and
Laurasian mainlands was often the basis for explaining
some peculiar features of the Cretaceous terrestrial verte-
brates found in the Periadriatic region and in other similar
domains. The faunas of the region have been described
as depauperate (Benton et al/., 1997), relict (Signore ef
al., 2001; Evans et al., 2004), endemic (Grigorescu et al.,
1999; Dal Sasso, 2001, 2004) and dwarf (Benton et al.,
1997; Jianu & Weishampel, 1999; Dalla Vecchia ef al.,
2000, 2001; Dal Sasso, 2001, 2004; Dalla Vecchia, 2002,
2009). The presence of dwarf and endemic faunas has
been considered as further evidence of life in confined,
isolated environment.

The taxon Compsognathidae is a clade of small coe-
lurosaurs presently known from the Late Jurassic and
Early Cretaceous. As one can see from the following list,
the oldest evidence in the fossil record is from the Late
Jurassic of Germany and, possibly, Portugal:

Juravenator starki - Kimmeridgian, Germany
(GGhlich & Chiappe, 2006)

Compsognathus longipes - Kimmeridgian, Germany;
Tithonian, France (Peyer, 2006)

compsognathid teeth - Kimmeridgian, Portugal
(Zinke, 1998)

cf. Aristosuchus - Berriasian, Romania
(Benton et al., 1997)

Aristosuchus pusillus - Barremian, England
(Naish et al., 2001)

Sinosauropteryx prima - Barremian/Aptian boundary-
Aptian, China (Xu & Norell, 2006)

Sinocalliopteryx gigas - ?Barremian/Aptian boundary,
China (Xu & Norell, 2006)

Huaxiagnathus orientalis - Barremian/Aptian boundary,
China (Xu & Norell, 2006)

Scipionyx samniticus - Albian, Italy
(Dal Sasso & Signore, 1998a)

Mirischia asymmetrica - ?Albian, Brazil
(Naish et al., 2004)

Scipionyx samniticus was found in an outcrop of Albi-
an age, a time when compsognathids were already differ-
entiated and widespread. Its basal position on the cladog-
ram suggests that Scipionyx could represent a relict clade
of basal compsognathids that perhaps retained primitive
features as a consequence of a separate evolutionary path.
Its generally plesiomorphic condition in respect to all oth-
er compsognathids, however, might be partly due to the
early ontogenetic stage of the only known individual (See
Ontogenetic Assessment, Phylogenetic Analysis).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Extensional or transform plate boundary

Continental shelves w_yw_yv Convergent plate boundary

Carbonate platforms

——> Plate motion vector

i Deep basins and oceans

R Pietraroja (Matese area)

Fig. 17 - Palaeogeographic map of the Periadriatic Domain (the present Central Mediterranean area) in the central-western Tethys,
during the Aptian (120 Mya). Abbreviations: ACP) Apenninic Carbonate Platform (Laziale-Abruzzese-Campana Platform); Ad) Adri-
atic-Dinaric Platform; Al) Algeria; AP) Apulian Platform; Au) Austroalpine; Ba) Bagnolo Platform; Br) Briangonnais; C1) Calabria;
Ga-Tr) Gavrovo-Tripolitsa Platform; Ib) Iberia; Im) Imerese Basin; Io-t) Ionian trough; ITO) Inner Tauride Ocean; Lb) Lombard
Basin; Li) Ligure-Piemontese Ocean; LM) Lagonegro-Molise Basin; NCA) Northern Calcareous Alps; PCPA) Panormide Carbon-
ate Platform; Pel) Pelagonian; Pi) Pindos trough; Ro) Rodophe; Sa) Saccense Pelagic Plateau; Sak) Sakarya; Si) Sicano Basin;
S-M) Serbia-Macedonia; Ti) Titsa; Tp) Trapanese Pelagic Plateau; Tu) Tunisia; U-M) Umbria-Marche Basin; Val) Valais Ocean.
(Modified after Turco et al., 2007; Cillari et al., 2009).

Fig. 17 - Carta paleogeografica del Dominio Periadriatico (l’attuale Mediterraneo centrale) nella Tetide centro-occidentale, durante
l’Aptiano (120 milioni di anni fa). Abbreviazioni: ACP) Piattaforma Carbonatica Appenninica (Laziale-Abruzzese-Campana);
Ad) Piattaforma Adriatico-Dinarica; Al) Algeria; AP) Piattaforma Apula; Au) Austroalpino; Ba) Piattaforma di Bagnolo; Br) Bri-
anzonese; C1) Calabria; Ga-Tr) Piattaforma di Gavrovo-Tripolitsa; Ib) Iberia; Im) Bacino Imerese; Io-t) Fossa Ionica; ITO) Oceano
Interno della Tauride ; Lb) Bacino Lombardo; Li) Oceano Ligure-Piemontese; LM) Bacino Lagonegro-Molisano; NCA) Alpi Calcaree
Settentrionali; PCPA) Piattaforma Carbonatica della Panormide; Pel) Pelagoniano; Pi) Fossa di Pindos; Ro) Rodope; Sa) Plateau
Pelagico Saccense; Sak) Sakarya; Si) Bacino Sicano; S-M) Serbia-Macedonia; Ti) Titsa; Tp) Plateau Pelagico Trapanese; Tu) Tunisia;
U-M) Bacino Umbro-Marchigiano; Val) Oceano Vallese. (Modificato da Turco e? al., 2007; Cillari e? al., 2009).

Studying the Pietraroja herpetofauna, Evans et al.
(2004) noted that the only known amphibian, the alban-
erpetontid Ce/tedens, had its closest relatives in Britain
(Berriasian, Purbeck Limestone) and Spain (Barremi-
an, Las Hoyas), and that Derasmosaurus is regarded
as “one of the last recorded rhynchocephalians from
Laurasia” (albeit that southern Italy is a Gondwanan
domain). Moreover, the presence of Fichstaettisaurus
would provide another archaic “Laurasian link” with
both Germany (Tithonian, Solnhofen) and Spain (Ber-

riasian, Montsec), suggesting that at Pietraroja “the re-
covered tetrapod fauna is primarily Laurasian in charac-
ter” (Evans et al., 2004).

Following Nicosia et al. (2007), we remark that that
statement points to a problematic connection northwards
to Laurasia, which, based on the present palaeogeographi-
cal knowledge, is quite unlikely since Jurassic times (see
above). Therefore, the conclusion of Evans et al. (2004)
is evidence of a palaeontological problem. As stated by
Xu (2010), “prevailing biogeographical hypotheses of

24 CRISTIANO DAL SASSO & SIMONE MAGANUCO

European isolation seem to be at least partly a product of
incomplete fossil sampling” (1.e., too scattered findings or
too few known outcrops), “a factor that has bedevilled bio-
geographical investigations of many dinosaur groups”. The
Lagerstatten are rare windows of preservation. Just for that
reason, by preserving the delicate remains of small-sized
taxa, which elsewhere would not have had the possibility
to fossilise in the given time, those exceptional palaeonto-
logical sites may induce to infer faunal similarity among
distant localities. Those taxa, as the Lagerstàtten suggest,
might instead have been more spread out. Like other small
terrestrial reptiles, compsognathid theropods occur in
Solnhofen, Pietraroja, Araripe and Liaoning not because
they lived only there, but because, having small delicate
skeletons, they had more chance of becoming preserved
in the fine sediments that produced those deposits. In fact,
from the list above it is inferred that during the Cretaceous,
the Compsognathidae might have reached a cosmopolitan
distribution: once they appeared, probably in the Middle-
Late Jurassic of western Laurasia (Germany, France, ?Por-
tugal), they remained there for a long time (England), and
contemporarily they spread up to the western limits of
Laurasia (China), moving also to Gondwana lands (Brasil)
and to their continental margins (Pietraroja).

Many recent finds — such as the ceratopsian Ajka-
ceratops in Hungary (Osi et al., 2010), a leptoceratop-
sid in Sweden (Lindgren ef a/., 2007), the alvarezsaurid
Albertonykus in Canada (Longrich & Currie, 2009), the
therizinosauroids Nothronychus and Falcarius in USA
(Kirkland & Wolfe, 2001; Kirkland e? a/., 2005), the cen-
trosaurine ceratopsid Sinoceratops in China (Xu et al.,
2010), and the reinterpreted Gondwanan, rather than en-
demic Australian, dinosaurs (Agnolin et al., 2010) — in-
dicate that the biogeography of a number of taxa was in

fact cosmopolitan rather than regional, and demonstrate
that any palaeobiogeographical inference must be made
with caution.

For all these reasons, we agree with Evans et al. (2004)
on the relevance of temporal divergence between the taxa
from Pietraroja and from other localities more than on
their palaeogeographical meaning. For example, the Ital-
ian Eichstaettisaurus (E. gouldi) undoubtedly provides a
significant extension to the temporal range of Eichstaetti-
saurus, given that it is separated from the Solnhofen Eich-
staettisaurus (E. schroederi) by a temporal gap of more
than 40 million years (Gradstein ef a/., 2004) and that the
morphological differences with £. schroederi are minor
and do not justify distinction at the generic level (Evans
et al., 2004). Although the persistence of fauna under iso-
lating conditions for more than 40 million years seems
improbable at best, we agree with Evans et a/. (2004) that
the Pietraroja assemblage may represent a relictual fauna.
The central-western Tethys archipelago may explain the
relatively archaic character of Early Cretaceous lepido-
saurian assemblages, regardless of their ancestors having
been Laurasian or Gondwanian, or even cosmopolitan.

Therefore, Scipionyx samniticus might have been either
an endemic species or simply a more widely spread, Late
Jurassic-Early Cretaceous compsognathid stock that had
become geographically isolated and relict in the Apennine
platform, and maybe in other Periadriatic platforms, too.
Insular dwarfism, as previously suggested for Scipionyx
(Dal Sasso, 2001, 2004) and observed in other dinosaurs
(Benton et al., 1997; Jianu & Weishampel, 1999; Dalla
Vecchia et al., 2000, 2001; Dalla Vecchia, 2002, 2009;
Sander et al., 2006) cannot be ruled out. However, given
the averagely small size of compsognathids, this seems an
unnecessary survival strategy for the taxon.

MATERIAL

The holotype and only known specimen of Scipio-
nyx samniticus is housed at the Soprintendenza per i
Beni Archeologici di Salerno, Avellino, Benevento e
Caserta with the inventory number SBA-SA 163760.
The original slab of fine-grained, grey-yellowish
limestone, which at the moment of its find embedded
the dinosaur, was collected in at least three broken

pieces. The original adjoining slabs, which probably
preserved the distal elements of the hindlimbs and
the remaining tail, were unfortunately not collected.
The specimen was subsequently completed with other
slabs (Figs. 9, 10) collected from the same locality,
but originating from different layers (Todesco, pers.
comm., 1993).

METHODS

Preparation

In 1993, the fossil appeared lying on a composite
slab covered by a layer of sticky vinylic glue (Fig.
18). Just after collecting the specimen, Todesco (pers.
comm., 1993) had prepared it roughly with chisels un-
der the naked eye. The abundant glue and some un-
removed matrix prevented clear observation of the
anatomy of the specimen (Fig. 19A). The specimen
also displayed an unusually short, non-segmented and
seemingly stiffened tail. This tail was actually an arte-
fact created with polyester resin by the collector (To-
desco, pers. comm., 1993).

Restoration and complete preparation of the fragile
bones and delicate internal organs involved removal of the

glue and consolidation with proper resin (Paraloid B72).
Removal of the false tail revealed a few real tail vertebrae
underneath the “connection” point. As mentioned above,
these interventions were performed in Salerno between
1994 and 1997, but always entrusted to the Laboratory of
Palaeontology of the MSNM.

Scipionyx was properly prepared using only manual
mechanical techniques under an optical stereomicroscope
(magnification of 10 to 60x). Very small chisels and then
increasingly finer steel needles were employed (Fig. 20).
In order to avoid damage to, and contamination of, the
organic remains, acid preparation was never used. Despite
very careful preparation, no integumentary structure, such
as scales or protofeathers, were discovered (Dal Sasso,
pers. obs., 1993-2010).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY DIS

Fig. 18 - The Pietraroja theropod in the spring of 1993. Note the artificial tail, nestled in a groove to mimic excavation, and connected
to the last preserved caudal vertebra (arrow).
Fig. 18 - Il teropode di Pietraroja nella primavera del 1993. Si noti la coda artificiale, inserita in un solco scavato per simularne
l’estrazione, e collegata all’ultima vera vertebra caudale conservata (freccia).

Optical microscopy

To observe the anatomical features in this small-sized
specimen, we used a Leica MZ9-5 stereomicroscope
equipped with a plan 1.0x lens, 10/20x oculars, and a 0.63
to 6.0 zoom. A Wild Heerbrugg TYP 308700 camera lu-
cida was mounted on top for detailed drawings.

Photographs and drawings

The photographs published in this monograph were
taken at different times (1993-2010) with both analog and
digital cameras. A camera was mounted on top of the ste-
reomicroscope for close-ups in visible light. Photographs
under ultraviolet (UV) light were taken with a digital
SLR camera equipped with a UV Hoya filter. Drawings
of detailed views were created using a camera lucida; all
other drawings were based on printed photographs taken
under different (usually opposite) oblique light. Most il-
lustrations were drawn by Marco Auditore, who worked
closely with the authors. A complete list of authors of the
published photographs and drawings can be found on the
title page verso (Illustration Credits).

In all of the original line drawings of Scipionyx pub-
lished herein, bold lines indicate the visible limits of a

given element, thin lines indicate anatomical structures
within a given element, and hatched lines indicate the es-
timated limits of an element overlain by others.

Measurements

Measurements were taken with a digital caliper and
a goniometer; micro-measurements were obtained with
an optical micrometer. If not specified, length of a given
complete/fragmentary element indicates its maximum
length, and its height or width or diameter were taken per-
pendicular to the maximum length. If not specified, di-
ameter refers to both craniocaudal or mediolateral meas-
urements, or to any intermediate plane, according to the
exposed view of the given element. Measurements and
ratios cited in the text are grouped in Tables 1-3.

UV light analysis

Fossils from some Mesozoic Lagerstàtten (Solnhofen,
Liaoning) can reveal exceptional morphological details
when observed under UV light. Delicate bones, tiny su-
tures and, in particular, remains of soft body parts that are
often poorly discernible or cannot be seen in visible light,

26 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 19 - The skull of Scipionyx samniticus before (A) and after (B) final
preparation. Note the U-shaped frontoparietal gap (red arrow), formerly
interpreted as damage occurring prior to professional preparation, but
actually untouched by the collector’s home-made tools (black arrows,
scratches made by the tools).

Fig. 19 - Il cranio di Scipionyx samniticus, prima (A) e dopo la pre-
parazione finale (B). Si noti lo spazio frontoparietale (freccia rossa),
inizialmente imputato ad un danneggiamento precedente alla prepara-
zione finale e invece non toccato dagli strumenti artigianali dello sco-
pritore (solchi indicati dalle frecce nere).

fluoresce conspicuously under filtered UV (e.g. Tischlin-
ger, 2002; Frey et al., 2003; Hone er al., 2010). This tech-
nique can be used to show otherwise hidden bony sutures,
and to distinguish bones or soft parts from the underlying
matrix and from each other.

When examining and photographing Scipionyx sam-
niticus under UV light, we obtained very similar and
highly informative results. Under black light blue lamps
(Philips TL 8W/08 F8 T5/BLB), the bones of Scipionyx
are opaque brown and the soft tissues fluoresce in col-
ours varying from white-gold to blue-indigo. According
to previous studies (e.g., Hone et a/., 2010), the variety of
colours is caused by a combination of different absorption
and reflectance of the various minerals that comprise the
specimen and the surrounding matrix. For instance, bones
(notably excluding their epiphyses; see Appendicular Ar-
ticular Cartilages) appear relatively uniform in colour and
reflectance, indicating homogeneous preservation; patch-
es of most soft tissues are seen as a bright fluorescence,
suggesting that they are phosphatised (Hone et a/., 2010).
Other tissues, such as the liver remains, fluoresce darkly
because of pigment remains (Ruben ef a/., 1999). Glue,
as well as restored parts and other artefacts made by syn-
thetic materials, fluoresce light blue.

Fig. 20 - The restoration and complete preparation of the tiny Scipio-
nyx and its delicate internal organs, done entirely under the optical ste-
reomicroscope with manual mechanical techniques, required about 300
hours.

Fig. 20 - Il restauro e la preparazione completa del piccolo Scipionyx
e dei suoi organi interni sono stati condotti sempre sotto uno stereomi-
croscopio ottico, con tecniche meccaniche manuali, e hanno richiesto
circa 300 ore di lavoro.

CT scan analysis

The specimen was analysed by X-ray computed to-
mography (CT) at the Radiology Department, Ospedale
Maggiore, Milan. The equipment used consisted in a Sie-
mens Somatom Definition Dual Source CT Scanner. The
best CT imaging was obtained with a bone algorithm on
transverse (axial) slices, with scan parameters 120 kV, 120
mA, and slice thickness of 0.3 mm. Data was exported in
DICOM format using eFilm (v. 1.5.3, Merge eFilm, To-
ronto). Analysis and post-processing were performed by
Armando Cioffi, Siemens, Milan. In general, CT scanning
did not provide high resolution images of the tiny bones
of Scipionyx. The extreme mediolateral diagenetic flat-
tening of the specimen was a complicating factor. Nev-
ertheless, this analysis was crucial in the understanding
of some obscured features of the skull, the path of the
intestinal loops and, especially, the proximal shape of the
ischia, which are intimately sandwiched in between the
two femoral shafts.

SEM analysis

Scanning electron microscopy (SEM) images and el-
emental peaks were obtained by analysing microsamples
with a Jeol JSM 5610 LV (IXRF Systems Inc.) equipped

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 27

with an EDS 500 spectrometer. The instruments are prop-
erty of the MSNM, and are housed in the Electron Micro-
scope Laboratory of our institute. The microsamples, no
more than 0.5-0.7 mm in diameter each, were taken from
carefully selected areas of the specimen (see Appendix
7) with the following procedure. The microsamples were
obtained mechanically through shallow incisions made
with a tiny widia blade. Uplift, rather than a cutting mo-
tion, was used in order to favour detachment running as
much as possible along the natural lines of separation of
the biological structures (for example, the endomysium-
sarcolemma of the muscle fibres). A steel point dipped in
Plasticine was used to pick up and drop the microsamples
immediately onto a double-sided carbon tape adhering to
the centre of the SEM sample holders (pin stubs). Most
of the microsamples were fixed in a reversed position
on the double-sided tape surface in order to expose their
“naturally” detached surfaces. This is crucial not only for
observing biological elements completely void of arte-
facts, but also for the element microanalysis. As the spec-
trometer of the SEM microprobe can analyse only surface
layers, it is Important to examine the internal surface of
freshly fragmented samples to minimise the risk of de-
tecting exogenous contaminants (Schweitzer ef a/., 2008).
Initial screening was performed on uncoated specimens;
subsequently, the samples were gold-coated in order to
enhance their conductivity to the electrons beamed by the
SEM, which produced images with a definitely higher
definition. Both SEM analysis and post-processing were
performed by Michele Zilioli (MSNM).

Anatomical terms

We refer to the margins and surfaces of the bones ac-
cording to their orientation in vivo, independently of the
position they assumed during fossilisation. Following
Weishampel et al. (2004), we adopt the terminology of
the Nomina Anatomica Veterinaria (AAVV, 2005) and the
Nomina Anatomica Avium (Baumel et al., 1993): ventral
(towards the belly), dorsal (towards the back), cranial
(towards the head), caudal (towards the tail), medial (to-
wards the midline) and lateral (away from the midline).
Thus, lateral can also indicate surfaces or structures inter-
nal to the bone, exposed by the removal of other, covering
bony layers.

In describing head elements, the term cranial was re-
placed by rostral (towards the tip, or rostrum, of the head),
the term cranial having no meaning with respect to the
head itself. Proximal (towards the mass of the body) and
distal (away from the mass of the body) are used to desig-
nate appendages, like segments or elements of a limb, and
also regions of the tail (e.g., proximal, middle and distal
caudal vertebrae). For the manus, palmar is used to desig-
nate the surface directed towards the ground, and dorsal is
used for the opposite surface.

Concerning the orientation of the teeth within the
Jaws, we followed the dental nomenclature specified by
Edmund (1969; see also Smith & Dodson, 2003): mesial
(the edge of a tooth towards the symphysis or premaxil-
lary midline), distal (the edge away from the symphysis
along the tooth row), labial (the surface of a tooth towards
the lip or the cheek) and lingual (the surface towards the
tongue).

Finally, for vertebrae and ribs we adopted the
terms specified by O’Connor (2007), and for vertebral
laminae and fossae we followed Wilson (1999); for
the gastralia we adopted the terminology of Claessens
(2004), whereas we followed Carrano & Hutchinson
(2002) in the nomenclature of pelvic and hindlimb
muscles. The terminology we used for all other soft
tissues is the one that the authors mentioned in the
text use in describing homologous elements in vivo in
extant vertebrates.

Systematic terms
The systematic terms, if not differently indicated, are
referred to Senter (2007) and Holtz et al. (2004).
Anatomical abbreviations
AIl anatomical abbreviations used in the text and in
the illustrations are listed in alphabetical order in the cov-

er flaps (small font size) and in Appendix 1 (normal font
size).

Institutional abbreviations

BSP Bayerische Staatssammlung fiir Palaontologie
und historische Geologie, Miinchen,
Deutschland

CEUM College of Eastern Utah Museum, Price, USA

DINO Dinosaur National Monument, Jensen, USA

FIP Florida Institute of Paleontology, Dania
Beach, USA

FMNH Field Museum of Natural History, Chicago,
USA

IPFUB Institut fiir Geologische Wissenschaften der

FU Berlin, Fachbereich Palàontologie, Berlin,
Deutschland

IVPP Institute of Vertebrate Paleontology and
Paleoanthropology, Beijing, China

MGI Mongolian Geological Institute, Ulaanbaatar,
Mongolia

ML Museu da Lourinhà, Portugal

MNHN Muséun National d’Histoire Naturelle, Paris,

France
MPN Museo di Paleontologia, Napoli, Italia

MSNM Museo di Storia Naturale di Milano, Milano,

Italia

NIGP Nanjing Institute of Geology and
Paleontology, Nanjing, China

NMC Canadian Museum of Nature, Ottawa, Canada
RTMP

Royal Tyrrell Museum of Palaeontology,
Drumheller, Canada

SBA-SA Soprintendenza per i Beni Archeologici di
Salerno Avellino Benevento e
Caserta, Salerno, Italia

O & SIMONE MAGANUC

- Overall view of the holotype (and only known specimen) of Scipionyx samniticus. Scale bar = 20 mm.
- Vista generale dell’olotipo (e unico esemplare conosciuto) di Scipionyx samniticus. Scala metrica = 20 mm.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Fig. 22 - Synoptic table of the osteology of Scipionyx samniticus (titles of chapters/subchapters for labels, related figures given in
brackets). Soft tissues have been omitted from this line drawing.

Fig. 22 - Quadro d’unione dell’osteologia di Scipionyx samniticus (nomenclatura conforme ai titoli dei capitoli/sottocapitoli, figure
correlate fra parentesi). In questo disegno al tratto sono stati omessi i tessuti molli.

CRISTIANO DAL SASSO & SIMONE MAGANUCO

- Skull and mandible of Scipionyx samniticus. Scale bar = 10 mm.
ranio e mandibola di Scipionyx samniticus. Scala metrica = 10 mm.

S
C

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Fig. 24 - Line drawing of the bones illustrated in Fig. 23. Black areas indicate the matrix. See Appendix 1 or cover flaps for ab-
breviations.

Fig. 24 - Disegno al tratto delle ossa illustrate in Fig. 23. Lo sfondo nero indica la matrice. Vedi Appendice 1 o risvolti di copertina
per le abbreviazioni.

CRISTIANO DAL SASSO & SIMONE MAGANUCO

premaxilla i Di angular
prefrontal (i palatine i surangular

ectopterygoid | prearticular Li splenial

pterygoid || interdental plate epipterygoid
lacrimal articular exoccipital

orbitosphenoid dentary squamosal

Fig. 25 - Cranial and mandibular bones of Scipionyx samniticus. A) right outer elements of the skull; B) scleral plates and right hemi-
mandible. / Ossa del cranio e della mandibola di Scipionyx samniticus. A) elementi esterni del lato destro del cranio; B) placche sclerali
ed emimandibola destra.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

jugal i basioccipital / basisphenoid frontal
quadrate tooth supraoccipital
quadratojugal maxilla [i scleral plate

basisphenoid / parasphenoid ERE indeterminate bone

prootic postorbital

laterosphenoid parietal matrix

Fig. 25 - Cranial and mandibular bones of Scipionyx samniticus. C) right and unpaired inner elements of the skull; D) left elements and
indeterminate bones. / Ossa del cranio e della mandibola di Scipionyx samniticus. C) elementi interni del lato destro, ed elementi impari
del cranio; D) elementi del lato sinistro e ossa indeterminate.

34 CRISTIANO DAL SASSO & SIMONE MAGANUCO

PART I - OSTEOLOGY

Systematic Palaeontology

DINOSAURIA Owen, 1842
SAURISCHIA Seeley, 1887
THEROPODA Marsh, 1881
TETANURAE Gauthier, 1986
COELUROSAURIA von Huene, 1914
COMPSOGNATHIDAE Cope, 1871
SCIPIONYX SAMNITICUS Dal Sasso & Signore, 1998

Type and only species - Scipionyx samniticus Dal
Sasso & Signore, 1998.

Etymology (Dal Sasso & Signore, 1998a) - “Scipio”,
Latin male name, dedicated to Scipione Breislak, the ge-
ologist who first described the Pietraroja Plattenkalk, and
Publius Cornelius Scipio (nicknamed Africanus), consul
militaris of the Roman Army, who fought in the Mediter-
ranean area; and “onyx” (6vvé), Greek for “claw”. “Sam-
niticus”, Latin for “of the Samnium”, the ancient name
of the region that includes Pietraroja and the Benevento
Province. Scipionyx should be pronounced “ship-e-ònyx”.

Holotype - Nearly complete, articulated skeleton with
fossilised soft tissues, housed in Salerno (Campania), Ita-
ly, at the Soprintendenza per i Beni Archeologici di Saler-
no, Avellino, Benevento e Caserta with inventory number
SBA-SA 163760.

Type locality - Le Cavere quarry, Pietraroja, Bene-
vento Province (Campania), Italy: IGM (Istituto Geo-
grafico Militare) map sheet 162, III SW-Cusano Mutri,
41220352°18%N314332i53335F:

Type horizon and age - “Calcari selciferi e ittiolitiferi
di Pietraroja” Fm. (sensu Catenacci & Manfredini, 1963),
upper Plattenkalk horizon (sensu Carannante et a/., 2006).
Lower Cretaceous, Lower Albian.

Emended diagnosis - Compsognathid theropod with
five premaxillary teeth; sinusoidal ridge of supratempo-
ral fossa at frontoparietal contact; descending process of
squamosal distally squared; carpus composed by only
two stacked, well-ossified bones (radiale and distal carpal
1+2); distal carpal 1+2 lenticular, compressed proximo-
distally, non-semilunate, and fused, lacking any trace of
suture; manual digit III markedly longer (123%) than digit
I; cranial concavity of the preacetabular blade of the ilium
in lateral view facing cranially and slightly developed;
prominent ischial obturator process squared distally.

Remarks - Besides autapomorphic characters, Sci-
pionyx shows also a mix of characters, some shared with
other compsognathids and/or basal coelurosaurs (see
Description for detailed comparisons), and some poten-
tially autapomorphic that might be subject to ontoge-
netic variation. The elongated dentary, prolonged more
caudally than the maxilla, might represent a synapo-
morphic feature (Currie, pers. obs., 1999) shared with
Juravenator. A related condition, potentially diagnostic
and possibly autapomorphic, is the lower tooth row that
extends farther back than the upper tooth row. However,
given that: (1) the specimen is immature; (2) both max-
illa and dentary might have undergone significant elon-
gation during ontogeny, but it is not clear if they were
subjected to positive or negative allometry each with re-
spect to the other; and (3) this probable elongation pos-
sibly gave rise to new tooth positions, these characters
must be considered with caution until a mature speci-
men of Scipionyx is available. The lack of contact be-
tween the quadratojugal and the squamosal might be an
autapomorphy of Scipionyx among basal coelurosaurs.
The absence of an external mandibular fenestra is shared
with other compsognathids and a few other basal coe-
lurosaurs. Similar to many compsognathids, Scipionyx
has dorsal neural spines that are relatively low and sig-
nificantly expanded craniocaudally, in particular at the
mid-top/top, with craniocaudally flat apical margins and
beak-like extensions on the cranial and caudal margin of
the neural spine, just below the apical margin. In Scipio-
nyx, the tip of the ungual phalanx of digit I ends at the
level of the distal end of phalanx 1 of digit II, a condi-
tion indicated as diagnostic of Compsognathus by Peyer
(2006). According to this author and to our phylogeny,
other compsognathid synapomorphies present in Scipio-
nyx are shaft diameter of phalanx I-1 equal/greater than
shaft diameter of the radius; manual unguals, especially
those of digits II and III, that are weakly curved, short
and wide; only slightly inclined dorsal transverse proc-
esses; absence of pleurocoels in dorsal vertebrae; and the
presence of a proportionally large skull (see Ontogenetic
Assessment). Lastly, Scipionyx shares with many other
compsognathids unserrated rostralmost but serrated lat-
eral teeth, hair-like cervical ribs, an orthopubic pubis,
and a pubic foot that lacks a cranial process but has a
well-developed caudal process.

OSTEOLOGICAL DESCRIPTION AND COMPARISONS

General features

The skeleton of Scipionyx lies on its left side, in near-
ly complete anatomical articulation (Figs. 21, 22). The
specimen measures 237 mm from the tip of the premax-
illa to the last (9) preserved caudal vertebra. The hind-
limbs of Scipionyx are missing distal to the proximal
epipodials, as are most of the tail and the right manual
claw II. The in situ articulation of the skeleton suggests
that the absence of these elements cannot be explained
by traumatic pre mortem events, nor by taphonomical
dispersal conditions (see Skeletal Taphonomy): they

were simply not collected (see Material). The head is up-
turned relative to the neck, but is not in an opisthotonic
condition (see Ostrom, 1978); the jaws are open. Some
bones and morphological structures are not exposed, but
their outlines are still visible as reliefs where they over-
lap each other (e.g., left humerus, right paraoccipital
process — see hatched lines in figures).

Taxonomic comparisons are made with other comp-
sognathids, basal coelurosaurians and basal forms of
more advanced theropods, in order to provide a detailed
description useful in elucidating phylogenetic relation-
ships and evolutionary trajectories within Coelurosauria.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 3 5

SKULL AND MANDIBLE

Scipionyx samniticus has a very large skull — meas-
uring half the length of the presacral vertebral column —
with well-developed smooth frontal and nasal bones and a
sloping postorbital region; the snout is short and triangu-
lar; the orbits are large and rounded, measuring a third of
the skull length; and the tooth row consists of sharp, fang-
like teeth (Figs. 23-25). No foramina for cranial nerves
are visible on the exposed parts of the skull bones.

Most skull bones are preserved in anatomical connec-
tion; during diagenesis, the bones of the left side have
been shifted 2 mm ventrally with respect to those of the
right side, as can be clearly seen in the cranial vault, in the
maxillae and in the mandible. This fact has helped inter-
pret some uncertain structures.

The comparatively large skull of Scipionyx matches the
general compsognathid bauplan, although there are some
proportions — most likely linked to the ontogenetic stage
of the specimen — that at first glance are reminiscent of
birds and ornithomimosaurs. The snout of Scipionyx, for
example, is comparatively shorter than in most non-avian
theropods (Padian, 2004). In the caudal half of the skull,
the sloping caudal skull table, and the arrangement and the
proportions of the bones as well, are somewhat reminiscent
of some ornithomimosaurs (see comparisons below), such
as Garudimimus (Kobayashi & Barsbold, 2005). These
similarities, especially those that appear birdlike, are prob-
ably accentuated by the immaturity of the individual, as is
the case for Bambiraptor (Burnham et al., 2000). Cranial
similarities and differences between Scipionyx and other

theropods are discussed in the following description of the
skull. The characters that can be linked to its ontogenetic
stage, such as size and proportions of the skull, are looked
at in more detail in the Ontogenetic Assessment section.

Cranial openings (Fig. 26)

Apertura nasi ossea - In Scipionyx, the bony nasal
opening (apertura nasi ossea in formal nomenclature, in-
appropriately called the “external naris” or simply “naris”
in palaeontology) is a drop-shaped opening, rostrocaudal-
ly elongated (although not as elongated as in Sinosaurop-
teryx and Huaxiagnathus), relatively large (as long as the
premaxilla) and extending caudally beyond the rostralmost
margin of the antorbital fossa. Its inner opening (called
the “naris” by most paleontologists) opens rostrally at the
level of the distal edge of the third premaxillary tooth and
terminates caudally at the level of the distal edge of m2.
Rostral to the inner opening, the outer opening (i.e., the
rim of the narial fossa) is visible as a depression on the
body of the premaxilla, and its mid-caudal outline coin-
cides with that of the inner opening.

The bony nasal opening is clasped in equal halves
by the premaxilla (rostrally) and by the nasal (caudally),
whose thin subnarial processes prevent the maxilla from
bordering the opening. The maxilla is excluded from the
opening also in other compsognathids and in other coeluro-
saurs (e.g., Caudipteryx, Incisivosaurus, Archaeopteryx),

Fig. 26 - Cranial openings of Scipionyx samniticus. Black areas indicate the matrix. See Appendix lor cover flaps for abbreviations.
Fig. 26 - Aperture del cranio di Scipionyx samniticus. Lo sfondo nero indica la matrice. Vedi Appendice 1 o risvolti di copertina per
le abbreviazioni.

36 CRISTIANO DAL SASSO & SIMONE MAGANUCO

although the contact between the premaxilla and the nasal
occurs differently in each taxon (see Premaxilla). The in-
ner bony nasal opening opens rostrally at the mid-level
of the third premaxillary crown in both Scipionyx and
Compsognathus (Peyer, 2006; Ostrom, 1978). Therefore,
the rostral margin of the nasal opening in these taxa is
caudally retracted with respect to that of Sinosauropteryx,
Sinocalliopteryx, Huaxiagnathus, Juravenator, Procera-
tosaurus, Dilong and Guanlong, in which the bone rostral
to the opening is as thin as the internarial bar.

Scipionyx is the only compsognathid with partial over-
lapping between the nasal opening and the antorbital fos-
sa. In Dilong, Ornitholestes (Osborn, 1916: fig. 1; Rauhut,
2003: fig. SH) and, possibly, Huaxiagnathus (Hwang et
al., 2004: fig. 2), the caudal margin terminates where the
rostral margin of the antorbital fossa begins. Other coe-
lurosaurs, such as Proceratosaurus (Woodward, 1910),
Guanlong (Xu et al., 2006), Incisivosaurus (Xu et al.,
2002a), Sinovenator (Xu et al., 2002b) and Archaeopteryx
(Elzanowski, 2002), are more similar to Scipionyx in hav-
ing a large bony nasal opening extending caudally beyond
the rostral margin of the antorbital fossa. Both conditions
are present in birds (Padian, 2004). This suggests that the
condition in Scipionyx is not necessarily a consequence of
the shortness of the snout due to the ontogenetic stage of
the specimen (see Ontogenetic Assessment).

Through the right nasal opening of Scipionyx, the su-
pranarial process of the left premaxilla and the medioven-
tral surface of the left ?nasal are visible.

Antorbital fossa - This is a marked, subtriangular de-
pression that includes the antorbital, maxillary and pro-
maxillary fenestrae. The antorbital fossa is surrounded
mainly by the maxilla (rostrally and ventrally) and by the
lacrimal (caudally and dorsally), with small contributions
from the jugal (caudoventral corner) and the nasal (dorso-
rostral margin).

Antorbital fenestra - A D-shaped antorbital fenestra
occupies most of the antorbital fossa, again with the par-
ticipation of the lacrimal and the jugal; the remaining
margins are entirely formed by the maxillary medial wall
of the fossa.

The antorbital fenestra is subequal in length to the or-
bit in Juravenator, Compsognathus and Sinocalliopteryx,
shorter than the orbit in Sinosauropteryx, and even shorter
in Scipionyx (see Ontogenetic Assessment). Thus, the sit-
uation in Scipionyx is somewhat reminiscent of Archae-
opteryx (Elzanowski, 2002), birds (Padian, 2004) and the
basal oviraptorosaurs Caudipteryx (Ji et al., 1998) and
Incisivosaurus (Xu et al., 2002a). An antorbital fenestra
that is shorter than the orbit is also present in some dei-
nonychosaurs, such as Velociraptor (Sues, 1977), Sinor-
nithosaurus (Xu & Wu, 2001) and the basal troodontid
Jinfengopteryx (Ji et al., 2005); therefore, this is definitely
a widespread condition within the Coelurosauria.

Maxillary fenestrae - The maxillary and the promax-
illary fenestrae are situated at the rostral border of the an-
torbital fossa, with the former dorsal to the latter. They are
separated from each other by a promaxillary strut (Fig.
27). Both fenestrae open and face laterally within the
maxillary medial wall. Their rostral borders are delimited
by the ascending process of the maxilla. The promaxillary

fenestra is small and oval in shape. Its rostral extension
may be slightly concealed in its lateral view by the as-
cending process of the maxilla. The maxillary fenestra has
the shape of a subrectangular trapezium. Underlying ma-
trix is visible through the promaxillary fenestra, while the
maxillary fenestra has an osseous medial wall (Fig. 27).

The maxillary fenestra in Sinosauropteryx and Comp-
sognathus is as large as that in Scipionyx. In Juravenator
and Huaxiagnathus, it is smaller. The shape and position of
the maxillary fenestra varies in compsognathids: Jurave-
nator, Huaxiagnathus, Sinosauropteryx and Compsogna-
thus have a rounded maxillary fenestra that is completely
enclosed in the centre of the maxillary medial wall of the
antorbital fossa. With the possible exception of Sinosau-
ropteryx, the fenestra is perforated in these taxa (Currie &
Chen, 2001). The maxillary fenestra is perforated in most
other theropods, with the exception of some troodontids
(Makovicky et al., 2003; Makovicky & Norell, 2004), in
which it is closed by a continuous medial wall.

A promaxillary fenestra is also present in most other
coelurosaurs, with the exception of most troodontids —
among which it was identified only in the basal form Sino-
venator (Xu et al., 2002b) — and, possibily, in some theriz-
inosauroids (Rauhut, 2003; but see Kundrat ef a/., 2008).
Peyer (2006) reported that Sinosauropteryx and Comp-
sognathus are the only compsognathids with an identifi-
able promaxillary fenestra. The promaxillary fenestra of
Scipionyx is comparable in size but differs in its outline:
it resembles more closely that of the basal tyrannosauroid
Dilong (Xu et al., 2004) in being small, located rostro-
ventral to the maxillary fenestra, and confined to the ros-
tral margin of the antorbital fossa. A similar arrangement

Fig. 27 - Close-up of the right antorbital region of Scipionyx samniticus.
See Appendix 1 or cover flaps for abbreviations. Scale bar = 1 mm.
Fig. 27 - Particolare della regione antorbitale destra di Scipionyx sam-
niticus. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.
Scala metrica = 1 mm.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 37

of the fenestrae, one dorsal to the other, is known for some
Ornithomimosauria (Rauhut, 2003: fig. 6E; Makovicky e?
al., 2004: fig 6.2B), although, according to Kobayashi &
Barsbold (2005), the dorsalmost opening in this taxon
would be the promaxillary fenestra.

Orbit - The orbit of Scipionyx is circular and larger
than the antorbital fossa. It is almost equally bordered by
the lacrimal (rostrally), the frontal (dorsally), the postor-
bital (caudodorsally and caudally) and the jugal (ventrally
and caudoventrally). A small contribution of the prefron-
tal (rostrodorsally) is also present.

A large and circular orbit is clearly present in the
compsognathids for which the orbital region has been
reconstructed, such as Compsognathus, Juravenator and
Sinosauropteryx, but also in many other adult theropods
that attained a small-to-medium body size, independently
of their phylogenetic affinities (e.g., Eoraptor, coelophys-
oids, basal tyrannosauroids, Ornitholestes, therizinosau-
roids, ormithomimosaurs, oviraptorosaurs, deinonycho-
saurs, alvarezsaurids, Avialae). The frontal and prefron-
tal contribution to the margin of the orbit in Scipionyx is
reminiscent of the condition in basal ornithomimosaurs
(Barsbold & Perle, 1984; Pérez-Moreno et al., 1994;
Kobayashi & Barsbold, 2005: 2A, D), some basal saur-
ischians like Eoraptor (Sereno et al., 1993), and in ther-
izinosauroids and Ornitholestes. A similar contribution to
the margin of the orbit occurs also in the alvarezsaurids
(Chiappe et al., 2002), but here the orbit flows caudally
into the infratemporal fenestra because of the lack of the
postorbital process of the jugal.

Supratemporal fossa - This is a subrectangular de-
pression that comprises the supratemporal fenestra. Usu-
ally facing dorsally, it is exposed laterally in Scipionyx
because of diagenetic crushing of the skull. It extends
rostrally up to the frontoparietal suture, and is well-sep-
arated from the controlateral fossa, as in Compsognathus
(Peyer, 2006). The dorsal portion of the parietals, which
separates left and right supratemporal fenestrae, is a broad
horizontal plate without any trace of a depression, unlike
in the majority of theropods (Weishampel et a/., 2004). As
a consequence, the medial margin of the supratemporal
fenestra coincides with that of the fossa.

Supratemporal fenestra - This opening is suboval
and bordered by the parietal (medially and rostromedial-
ly), the frontal or the postorbital (rostrally, see Postorbital
for further comments), the postorbital (laterally) and the
squamosal (caudolaterally and caudally).

Whereas the parietal borders most of the supratem-
poral fenestra in Scipionyx and Compsognathus (Peyer,
2006), in Juravenator it borders the fenestra only for a
short, medial tract so that the pre-eminent contribution
is made by the frontal (Gòhlich & Chiappe, 2006: fig.
2a). Three bones are visible through the supratemporal
fenestra: the prootic, the laterosphenoid and the ventral
portion of the parietal.

Infratemporal fenestra - This fenestra opens just
below the supratemporal fenestra. It is rectangular — the
postorbital bar and the ascending ramus of the quadrato-
jugal parallel the quadrate — with four corners formed by
the postorbital (rostrodorsal), squamosal (dorsocaudal),

quadratojugal (caudoventral) and jugal (ventrorostral).
The former hypothesis that the mid-height caudal con-
striction of this fenestra might be due to a slight rostral
displacement of the quadratojugal ramus of the squamo-
sal cannot be ruled out, but in all likelihood the orientation
of this ramus is natural. In fact, a comparable mid-height
constriction appears to be present, for example, in Sino-
sauropteryx (Currie & Chen, 2001: fig. 3b) and in Sau-
rornithoides (Makovicky & Norell, 2004: fig. 9.1), and is
also present in other basal tetanurans such as A//osaurus
and Monolophosaurus. In those taxa, as in Scipionyx, the
squamosal forms most of the protrusion, whereas in ty-
rannosaurids (Holtz, 2004), and possibly in Ornitholestes
(Osborn, 1916, fig.1; Rauhut, 2003, fig. 5H), the infratem-
poral flange is due to the synapomorphic extension of both
the squamosal and the quadratojugal.

Palatal fenestrae - Because of diagenetic distorsion
of the skull, the following palatal fenestrae have been
partially exposed: the rostromedial portion of the subtem-
poral fenestra, bordered by the ectopterygoid process of
the pterygoid, and visible along the caudoventral margin
of the orbit; the suborbital fenestra, exposed along the
ventrorostral margin of the orbit, bordered by the palatine
(rostrally), the pterygoid (medially) and the ectopterygoid
(mediocaudally), and reduced in size with respect to the
orbit; the subsidiary palatal fenestra, exposed only in its
caudal part, clasped by the pterygoid and partly bordered
by the pterygoid ramus of the palatine; the internal naris
(choana), partly visible in the centre of the antorbital fe-
nestra as a C-shaped opening, bordered by the vomeral
and the rostral maxillary processes of the dorsally dis-
placed left palatine; and the interpterygoid vacuity, delim-
ited by the palatal rami of the left and right pterygoid.

Peyer (2006) observed that in Compsognathus the
rostrolateral border of the palatine marked the caudal
border of the internal naris. No other details of the pal-
atal fenestrae have been described for compsognathids.
Some palatal bones and relative fenestrae are indeed vis-
ible in Sinocalliopteryx (Ji et al., 2007a) and especially
in Juravenator (Gòhlich & Chiappe, 2006), but they
were not described and their shape and margins have not
been reconstructed. A subsidiary palatal fenestra (sensu
Ostrom, 1969; “pterygopalatine fenestra” in Norell &
Makovicky, 2004) is present also in tyrannosaurids, or-
nithomimosaurs, deinonychosaurs (Rauhut, 2003) and in
therizinosaurids (Clark et a/., 1994).

Foramen magnum - The medialmost third of the
straight ventral margin of the right supraoccipital prob-
ably bordered the foramen magnum dorsally.

Scleral plates

The bony plates which formed the scleral rings (i.e.,
the bony reinforcements of the eye balls) occupy most of
the orbit of Scipionyx. Like many bones of the left side,
the left scleral ring has slid a couple of mm ventrally, so
that its dorsal arch is now visible at the centre of the orbit.
Despite this displacement, it seems to be complete, and
its curvature can be followed just below the palatal bones.
On the other hand, the right scleral ring is broken into at
least 4 arched fragments, each composed of 3 to 5 plates.

59 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 28 - The scleral rings of Scipionyx samniticus. A) ultraviolet-induced fluorescence highlights the overlapping of the plates.
B) close-up, under the optical microscope, of the dorsal arch of the left scleral ring. Scale bar = 1 mm.

Fig. 28 - Anelli sclerali di Scipionyx samniticus. A) fluorescenza indotta da luce ultravioletta che evidenzia le sovrapposizioni tra le
placche. B) particolare al microscopio ottico dell’arco dorsale dell’anello sclerale sinistro. Scala metrica = 1 mm.

The apparent difference in shape and size of the scle-
ral plates is due to different patterns of superimposition.
The plates narrow towards their somewhat irregular or
zigzagged contact margins, which are more clearly vis-
ible under UV light (Fig. 28A). The exact number of
plates that formed each scleral ring cannot be directly
determined, but a minimum number of 16 plates can be
assumed (each of the 4 fragments of right ring makes up
almost a quarter of the ring).

Among compsognathids, Sinocalliopteryx had at least
10 scleral elements, in the form of subrectangular to trap-
ezoidal plates (Ji et a/., 2007a: fig. 2a-b). Scleral rings are
recorded in the Sihetun Sinosauropteryx, but they are cov-
ered externally by a black substance, possibly represent-
ing the decayed retina or iris of the eye (Currie & Chen,
2001). Subrectangular scleral plates with thin, sharp edges
were observed in Sinornithosaurus (Xu & Wu, 2001). In
the holotype of Garudimimus, the left side of the skull
preserves 11 articulated scleral plates (Kobayashi & Bars-
bold, 2005). The scleral ring of Jinfengopteryx is report-
ed to consist of 12-13 plates (Ji et a/., 2005). 24 scleral
plates are figured in coelophysoids (Tykoski & Rowe,
2004: fig. 3.2), and 11-12 are reported in Archaeopteryx
(Elzanowski, 2002). In extant birds (Bellairs & Jenkins,
1960), the number of seleral ossicles varies, ranging from
10 to 18, with 14-15 being the most common numbers.

Dermal skull roof

Premaxilla - The right premaxilla of Scipionyx is al-
most as deep as it is long, tapering rostrally and terminat-
ing in a rounded rostral tip (Fig. 29). The rostralmost mar-
gin is vertical for only a very short tract and then curves
gradually in a dorsocaudal direction, continuing in a cau-
dally directed supranarial process inclined 60° respect to
the vertical plane. A similar condition is present in Jurave-
nator, Huaxiagnathus, Sinocalliopteryx and Compso-

gnathus (contra Peyer, 2006), but not in Sinosauropteryx,
where the tip of the snout is truncated, the rostral margin
of the premaxilla is vertical and the supranarial process is
inclined about 35° caudally to the vertical plane (Currie
& Chen, 2001: 2a).

The premaxilla of Scipionyx bears five teeth (see Den-
tition). Several nutritive foramina can be seen on its sur-
face. The premaxillary foramina are larger in Juravenator
(Gohlich & Chiappe, 2006: fig. 2a), smaller but still well
visible in Sinocalliopteryx (Ji et al., 2007a: fig. 2a), but
absent in Compsognathus (Ostrom, 1978; Peyer, 2006).

In Scipionyx, the main body of the premaxilla in the
subnarial region is as long rostrocaudally as it is high dor-
soventrally; thus, it is higher than in Sinosauropteryx. The
caudal extension of the supranarial and subnarial proc-
esses of Scipionyx is almost the same, but the former is
longer than the latter. As for most theropods, including
Compsognathus and Sinosauropteryx, the subnarial proc-
ess of Scipionyx is narrow and relatively short. The dorsal
margin of the subnarial process of the premaxilla borders
the external naris ventrally. It contacts the subnarial proc-
ess of the nasal in a dorsocaudally-to-rostroventrally di-
rected sloping suture, excluding the maxilla from the ex-
ternal nares. In Juravenator, however, the longest process
is the subnarial one, which forms the entire ventral margin
of the naris. In Sinocalliopteryx, the supranarial and sub-
narial rami are even longer, and in Huaxiagnathus, the
subnarial ramus reaches the level of the rostral margin
of the antorbital fossa. In Scipionyx, a tiny elliptical hole
that may be the subnarial foramen is present along the
premaxilla-maxilla suture, just in front of the lanceolate
tip of the subnarial process of the nasal (Figs. 27, 29).

The premaxillary-maxillary contact is almost vertical
in its ventral half and dorsocaudally directed in its dorsal
half. The contact between the two bones is more extensive
than in Juravenator, and their ventral margins close to
the suture are almost horizontal. The maxillary margin is
only slightly inclined rostrodorsally, and in any case does

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 39

Fig. 29 - Shaded drawings of the right nasal (A), maxilla (B) and premaxilla (C) of Scipionyx samniticus, and their sutural contacts.

See Appendix 1 or cover flaps for abbreviations.

Fig. 29 - Disegni ombreggiati di nasale (A), mascellare (B) e premascellare (C) destri di Scipionyx samniticus, e loro contatti suturali.

Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

not have the marked ventral notch visible in Juravenator
(Gohlich & Chiappe, 2006). Judging from the rostral half
of the supranarial process, the internarial bar of Scipionyx
is flattened dorsoventrally. The left premaxilla emerges in
medial view only with the supranarial process and a small
caudoventral portion.

Maxilla - The maxilla is a triangular bone, notched
caudally by the antorbital fenestra and pierced by two
smaller fenestrae — the maxillary and the promaxillary fe-
nestrae (Fig. 29). In lateral view, the maxilla seems to de-
velop in two different planes: a portion overhanging later-
ally, which delimits the antorbital fossa with two rami,
one horizontal bearing the tooth row, and one ascending
towards the nasal; and a portion lying on a more medial
plane, forming the medial wall of the antorbital fossa.

The passage between the two planes is well-marked, even
though the antorbital fossa is only slightly depressed be-
low the level of the surrounding bone.

The antorbital fossa is separated from the rest of max-
illa by a low, continuous but distinct ridge, as in Huaxia-
gnathus (Hwang et al., 2004) and in a few other theropods
(Rauhut, 2003). This ridge is rounded rostrally, then con-
tinues caudally ventral to the antorbital fenestra, running
parallel to the tooth row. The fossa deepens towards its ros-
tral rim, where it is pierced by the promaxillary fenestra.
Whereas the promaxillary fenestra appears as a hole, the
maxillary fenestra is not so apparent at first glance be-
cause it is paved by an osseous wall that is even more me-
dial than the maxillary medial wall. A similar wall in the
fenestra, belonging to the maxilla, was described in some
troodontids (Makovicky e? a/., 2003) and hypothesised to

40 CRISTIANO DAL SASSO & SIMONE MAGANUCO

be present also in Sinosauropteryx (Currie & Chen, 2001).
In Scipionyx, its belonging to the right maxilla, rather than
to an unidentified element of the left side of the skull, is
supported by the position of both left nasal and left max-
illa. These bones, which emerge in other areas, could have
invaded the background of the maxillary fenestra with
fragments, after breaking into pieces. However, the conti-
nuity of the bone that is seen within the fenestra rules out
this possibility.

The compression of the medial wall of the maxillary
in the rostral portion of the antorbital fossa has forced out
a longitudinal bar that would otherwise be visible only
in medial view. This bar is rostrocaudally directed and
slightly slanting with respect to the ventral margin of the
antorbital fossa. It arises in between the promaxillary and
maxillary fenestrae in the form of a promaxillary strut,
and terminates in correspondence to the ventral margin of
the antorbital fenestra. Ventral to this fenestra, the maxilla
gradually thins caudally and terminates in an acute triangle
which contacts the jugal with an elongated oblique suture.

Rostral to the antorbital fenestra, the ascending proc-
ess of the maxilla, thinner and lying in a lateral plane,
forms a long oblique suture with the subnarial process
of the nasal. As mentioned, the maxilla is excluded from
the caudal and ventral margin of the external naris by the
premaxillary-nasal contact, like in the vast majority of
theropods (e.g., Rauhut, 2003).

The dorsorostral margin of the maxilla is slightly con-
caved for a very brief tract Just below the premaxillary-
nasal contact (Chiappe, pers. obs., 2006), interrupting its
otherwise convex course from the ascending process to the
ventral margin. In our opinion, this concavity is homolo-
gous to the more evident one that in most teropods marks
the transition between the ascending ramus and the rostral
ramus. Therefore, contrary to Holtz et al. (2004) and Peyer
(2006), the rostral ramus is not lost in Scipionyx: the bone
possessing the primitive simple convex curve is strongly
reduced or, more probably, not yet fully developed on ac-
count of the immaturity of the specimen. Thus, the rostral
ramus of Scipionyx is poorly developed and shorter rostro-
caudally than it is high dorsoventrally. This is contrary to
what is seen in Juravenator and Compsognathus, and even
more so in Hwaxiagnathus and Sinocalliopteryx, which are
similar to the other coelurosaurs in having a maxillary ros-
tral ramus that is rostrocaudally as long as, or longer than,
it is long dorsoventrally. According to Holtz et al. (2004),
this condition is one of the diagnostic characters of gen-
eralised coelurosaurs, although it is present also in some
basal non-coelurosaurian tetanurans.

The antorbital fenestra is delimited rostrally by a por-
tion of the maxillary medial wall, the interfenestral bar
(pila interfenestralis). This bar separates the antorbital fe-
nestra from the maxillary fenestra. In Scipionyx, the max-
illary medial wall is not low and subparallel to the tooth
row as in Compsognathus (Peyer, 2006); rather, it gradu-
ally rises up rostrodorsally to form the interfenestral bar,
where it is tall half the height of the antorbital fossa, as is
often the case in theropods. As the dorsalmost portion of
the maxillary medial wall has been slightly covered by the
flattening of the nasal during diagenesis, the antorbital and
maxillary fenestrae seem to be bordered by the nasal, too.
The nasal possibly participates in forming the antorbital
fossa, but is excluded from the antorbital fenestra by the
sutural contact between the bifurcating caudal extremity

of the interfenestral bar of the maxilla and the horizontal
ramus of the lacrimal (see Nasal).

Above the dentigerous margin, the lateral surface of
the maxilla is pierced by numerous nutritive foramina, as
in Sinocalliopteryx and Sinosauropteryx (Currie & Chen,
2001). In contrast, the alveolar ramus in Compsognathus
(Ostrom, 1978; Peyer, 2006) is devoid of large nutrient
foramina. The right tooth row bears 7 teeth, with ml
resulting slightly procumbent because of the very faint
slope of the rostral portion of the tooth row towards the
premaxillary-maxillary contact; six teeth are preserved on
the left maxilla (see Dentition).

As it can be inferred by the position of its ventral mar-
gin, which terminates rostrally at the level of the right ml,
the left maxilla slid ventrally along the premaxillary-max-
illary suture more than the left premaxilla did, creating a
step between the two bones. Thanks to this sliding, the me-
dial surface of the left maxilla is exposed in two areas (Fig.
25D): inside the right antorbital fenestra, where the left
interfenestral bar is visible, paralleling the counterlateral
element; and below the dentigerous margin of the right
maxilla. In the latter area, along the medial side of the left
maxilla, at least 5 distinct interdental plates emerge, regu-
larly spaced, between maxillary teeth 2-7 (Fig. 30). They
are not fused to each other, so they do not form a continu-
ous lamina, nor do they seem fused to the maxilla. They
are rather large and bulky, sticking out with respect to the
medial margin of the maxilla, and change shape accord-
ing to the shape of the interdental gaps, the rostralmost
being subrectangular and the caudalmost subtriangular.
Large, triangular unfused interdental plates are present in
both German and French Compsognathus (Peyer, 2006).
Unfused interdental plates of variable shape but somewhat
comparable to those of Scipionyx, are described in the ba-
sal dromaeosaurid Sinornithosaurus (Xu & Wu, 2001).
The interdental plates of Archaeopteryx resemble those
of Scipionyx in being distinctly separated from the medial
margin of the maxillary bone and widely separated from
one another, although they change in shape in an opposite
manner, the rostral ones being subtriangular and the caudal
ones being subrectangular (Elzanowski, 2002). Unfused
interdental plates are present also in the Tyrannosauridae
(Currie ef al., 2003), but not in some basal tyrannosauroids
such as Tanycolagreus (Carpenter et al., 2005a).

Fig. 30 - Close-up of the dentigerous margin of the maxillae of Scipio-
nyx samniticus, showing teeth attachments and interdental plates. Scale
bar = 1 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 30 - Particolare del margine dentigero dei mascellari di Scipionyx
samniticus, in cui sono visibili le attaccature dei denti e le placche inter-
dentali. Scala metrica = 1 mm. Vedi Appendice 1 o risvolti di copertina
per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 41

Nasal - The nasal is a long and narrow wedge-shaped
plate of bone, with parallel medial and lateral margins and
a smooth surface. Because the skull was crushed, the right
nasal appears flattened, with lateral and dorsal surfaces in
one plane.

Rostrally, the nasal diverges to clasp the back of the
external naris. In addition, the supranarial process splits
rostrally clasping, in turn, the caudal tip of the premax-
illa. The subnarial process is hardly visible, on account of
its thinness and some parallel microfractures in the sur-
rounding bones, but appears to be long, very thin and with
a lanceolate process (Fig. 29). The subnarial process is
thin also in Compsognathus and in Juravenator (thinner
than reconstructed in Gohlich & Chiappe [2006: figs. 2a-
b]), although not as thin as in Scipionyx.

Dorsal to the maxillary fenestra, at mid-height of the
bone, the surface of the nasal is pierced by 4 small, oval,
rostrocaudally aligned, almost equidistant foramina (Figs.
27, 29). They might have served as passages for branches
of the superior nasal artery or led into internal pneumatic
chambers. A rostrolaterally directed row of at least 3 nu-
trient foramina is present in the nasal of Sinosauropteryx
(Currie & Chen, 2001: fig. 3d) just rostral to the vertical
ramus of the lacrimal. No foramina have been found in
Compsognathus (Peyer, 2006).

On account of the compression of the skull and the
consequent position of the nasal, the contacts between
the maxilla and the nasal, and their contribution to the
antorbital fossa, are not completely clear. Certainly, the
antorbital fossa does not invade the ventrolateral margin
of the nasal, contrary to the condition shown, for exam-
ple, by the allosauroids (Rauhut, 2003). It is difficult to
establish whether the nasal bordered the antorbital fossa
or whether it was excluded by a process of the maxilla.
The ascending process of the maxilla seems to have a
thin appendage that, running caudodorsally along the
dorsal margin of the maxillary fenestra, almost reaches
the interfenestral bar and might have continued caudally
up to the lacrimal.

Caudally, the nasals are so firmly sutured to the fron-
tals that Dal Sasso & Signore (1998a) were unable to
identify any nasofrontal suture. Photos at high magnifi-
cation (Fig. 31A-B) show an interdigitate type, roughly
W-shaped nasofrontal suture just dorsal to the lacrimal. In
dorsal view, beginning from the midline, this suture runs
rostrolaterally for a short way, then becomes transverse;
in its second half, it runs caudolaterally towards the pre-
frontal. A W-shaped nasofrontal suture at the level of the
vertical ramus of the lacrimal is found also in Compso-
gnathus (Peyer, 2006: fig. 4C), Sinocalliopteryx (Ji et al.,
1997a: fig. 2b), Juravenator (GG6hlich & Chiappe, 2006:
fig. 2a) and Huaxiagnathus (Hwang et al., 2004: fig 2A).
In Sinocalliopteryx the suture is an inverted W shape.

A thin chip of the left nasal is visible just dorsal to the
short exposure of the medial sagittal suture, and a larger
piece can be seen through the opening of the external
nares (Fig. 25D).

Lacrimal - The stout lacrimal, with a typical inverted
L-shape, is among the most robust cranial bones. Equal-
ly robust lacrimals can be found only in tyrannosaurids
and ornithomimosaurs.. Compsognathus, Juravenator,
Sinosauropteryx and Sinocalliopteryx have an inverted
L-shaped lacrimal like Scipionyx and the majority of

theropods, including basal coelurosaurians. The lacrimal
is T-shaped in the Deinonychosauria, Ornithomimosauria,
Oviraptorosauria and in Aves (Rauhut, 2003).

In Scipionyx, the vertical ramus is pillar-like, and its
base leans against a dorsal sulcus of the jugal. Dorsally, it
has a long, straight contact with the prefrontal, separating
the antorbital fenestra from the orbit. The caudal margin
of the vertical ramus is reinforced by a crest (very thin in
the ventral half of the bone and markedly expanded in the
dorsal half) which forms the rostral orbital margin.

The horizontal ramus runs rostrally, ventral to the base
of the prefrontal, to reach the ventral margin of the nasal
and the interfenestral bar of the maxilla. It is shorter than
the vertical ramus, a condition probably linked to the on-
togenetic stage of the individual (see Ontogenetic Assess-
ment). The horizontal ramus is reinforced by crests which
parallel the contact with the prefrontal. Among these
crests, the most marked one borders the lacrimal vacuity,
a depression located in the corner between the horizontal
and the vertical ramus. A smaller foramen opens laterally
at the angle between the two rami in Juravenator (Chi-
appe, pers. comm., 2006). Such a foramen is absent in
Scipionyx even under UV light (Fig. 31A, C).

As in most small-sized theropods, including other
compsognathids, there is no evidence of either a supraor-
bital crest or a lacrimal horn. In species that attained a me-
dium-to-large body size, these structures are often poorly
developed in young individuals, and become apparent
only in mature individuals (e.g., Madsen, 1976).

A part of the left horizontal ramus seen in medial view
and still contacting the extremity of the left maxillary
interfenestral bar is visible through the dorsalmost por-

Fig. 31 - Ultraviolet-induced fluorescence photograph (A), and close-
ups under visible light (B, C), documenting the position and shape of
the nasofrontal suture in Scipionyx samniticus (red arrow; black arrow
indicates a fracture). Note also the lacrimal vacuity in (A) and (C).
Scale bars = 1 mm. See Appendix 1 or cover flaps for abbreviations.
Fig. 31 - Fotografia in fluorescenza indotta da luce ultravioletta (A),
e particolari in luce visibile (B, C), che documentano la posizione e
l’andamento della sutura nasofrontale in Scipionyx samniticus (freccia
rossa; la freccia nera indica invece una frattura). Si noti anche la vacui-
tà lacrimale in (A) e (C). Scale metriche = 1 mm. Vedi Appendice 1 o
risvolti di copertina per le abbreviazioni.

42 CRISTIANO DAL SASSO & SIMONE MAGANUCO

tion of the antorbital fenestra (Fig. 25D). A piece of the
caudoventral portion of the jugal process is sandwiched
between the right palatine and the right lacrimal. The thin
bar of bone seen paralleling the rostral margin of the right
lacrimal at mid-shaft is possibly another fragment of the
same lacrimal.

Prefrontal - This is a triangular bone connecting the
lacrimal to the nasofrontal complex in close sutural con-
tact. In Scipionyx, the prefrontal is well-developed and
comparatively large and robust, surpassing the horizontal
ramus of the lacrimal in length and, therefore, preventing
any contact between the lacrimal and the frontal. In the
uncrushed skull, it must have been well-exposed also in
lateral view, expanding ventrally to terminate in a small
worm-like appendage that still contacts the caudomedial
face of the lacrimal. This appendage probably contributed
to the rostromedial margin of the orbit, but remained vis-
ible in lateral view.

A robust fragment of the left prefrontal is exposed in
medial view along the dorsorostral margin of the orbit.
It still contacts the ventral surface of the left frontal (see
below).

The prefrontal is greatly reduced in basal tetanuran
theropods (Holtz et al. 2004). A well-developed prefron-
tal, although less prominent than in Scipionyx, is reported
in Sinosauropteryx (Currie & Chen, 2001), Compsognath-
us (Peyer, 2006) and Sinocalliopteryx (Ji et al., 2007a),
possibly in Juravenator (Gòhlich & Chiappe, 2006: fig.
2a), and in a number of coelurosaurs (e.g., some tyranno-
sauroids, Ornitholestes). In therizinosauroids (e.g., Clark
et al., 1994; Kundrat et a/., 2008) and alvarezsaurids (Se-
reno, 2001; Chiappe et a/., 2002), as well as in some basal
coelophysoids, the prefrontal is much larger. The most
prominent prefrontals are reported for ornithomimosaurs
(e.g., Kobayashi & Lii, 2003: fig. 5B; Makovicky er al.,
2004; Kobayashi & Barsbold, 2005: 2a, d). In these ani-
mals, they continue ventrally, adhering along the caudo-
medial face of the lacrimal. Further comments on the size
of the prefrontal of Scipionyx are given in the Ontogenetic
Assessment section.

Postorbital - As in most theropods, the postorbital
of Scipionyx is triradiate in lateral view. The descending
ramus (jugal ramus) is the most developed one. On ac-
count of diagenetic crushing, even its contribution to the
caudomedial margin of the orbit can be seen. The jugal ra-
mus, however, is more limited in size than was previously
considered: in fact, the internal portion of the postorbital,
“ipo” in Dal Sasso & Signore (1998a: fig. 4), is reinter-
preted here as the right epipterygoid.

The supraorbital ramus is slightly shorter than the jugal
ramus, but equally contributes to the orbits. In Juravena-
tor, the jugal ramus of the postorbital is even longer than
in Scipionyx. This contrasts sharply with the condition
seen in Compsognathus (Peyer, 2006), where the supraor-
bital ramus is the longest and might have reached forward
over the orbit to meet the prefrontal, excluding the frontal
from the orbital margin. Looking only at the proportions
of these two rami, the postorbital of Scipionyx more close-
ly resembles that of ornithomimosaurs (e.g., Kobayashi &
Barsbold, 2005). No resemblence is seen with any basal
tetanurans, in which the supraorbital ramus is the shortest
process of the triradiate postorbital, nor with dromaeosau-

rids, in which the dorsal margin of the supraorbital proc-
ess is directed dorsomedially, not straight rostrally (Norell
& Makovicky, 2004).

The caudally directed squamosal ramus is the short-
est of the three rami. We cannot exclude the presence of
a fourth medial process, which might have contacted the
postorbital process of the parietal in the uncrushed skull,
excluding the frontal from the margin of the fenestra but
not from the fossa, as is the case in A//osaurus (Madsen,
1976).

Jugal - The jugal can be divided into the horizontal
body, a long flattened bar of bone forming the ventral
margin of the orbit and cheek, and the ascending process,
caudal to the orbit. The horizontal body terminates ros-
troventrally in the maxillary process, which forms an ob-
lique suture with the maxilla. Rostrodorsally, it presents a
horizontal sulcus accommodating the base of the lacrimal,
but it lacks any sublacrimal expansion, like in Compso-
gnathus (Peyer, 2006) and Juravenator (Gòhlich & Chi-
appe, 2006). Huaxiagnathus (Hwang et al., 2004: fig. 2a)
has a feeble expansion, whereas Sinocalliopteryx (Ji et
al., 2007a: fig. 2b) and most theropods (Rauhut, 2003)
have a clearly developed sublacrimal expansion.

As pointed out by Dal Sasso & Signore (1998a: fig.
4), the tip of the maxillary process reaches the caudoven-
tral corner of the antorbital fenestra like in many thero-
pods, including Compsognathus (Peyer, 2006) and, pos-
sibly, Juravenator (Gòhlich & Chiappe, 2006: fig. 2a).
Differently to Torvosaurus, Ornitholestes, allosauroids
and tyrannosauroids, the jugal of Scipionyx is devoid of
a crescentic depression marking a slight participation of
the antorbital fossa.

Caudally, and ventral to the infratemporal fenestra,
the horizontal body of the jugal reaches the quadratoju-
gal with a short process that appears bifid because of the
superimposition of the pointed apex of that bone. The
caudalmost portion is covered by the quadratojugal and,
thus, is hardly visible; rather than rod-like, it seems to be
“at least twice as tall dorsoventrally as it is wide trans-
versely”, as suggested by Peyer (2006). In fact, the dor-
soventral diameter of the quadratojugal process beneath
the infratemporal fenestra seems to be naturally twice as
long as its mediolateral diameter, the margin of the pro-
cess being rounded and regular, without any trace of dia-
genetic compression. The articulated skull of Scipionyx
shows a condition similar to that of the reconstructed
Compsognathus (Peyer, 2006: 4B). Detailed comparisons
with Sinosauropteryx can hardly be made, as the jugal
process was generically described as relatively short and
high (Currie & Chen, 2001).

The ascending process (postorbital process) is a thin
vertical bar that contacts the descending process of the
postorbital, separating the orbit from the infratemporal
fenestra. In Scipionyx, as well as in some other compsog-
nathids (Hwang et a/., 2004; Gòhlich & Chiappe, 2006;
Ji et al., 2007a), it seems to be more robust (i.e., thicker
rostrocaudally) than in the reconstructed skull of Comp-
sognathus (Peyer, 2006: fig. 4B).

Quadratojugal - The quadratojugal of Scipionyx is
L-shaped (or hook-shaped), with rami of equal length ta-
pering to simple pointed ends and forming an angle of
about 90°. No caudal process is present. The horizontal

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 43

Fig. 32 - Computed tomography image of Scipionyx samniticus, showing the bones of the left side. The positions of the left (red arrow)
and right quadrate (green arrow) reveal that the bump deforming the right quadratojugal is due to the medial condyle of the left quad-
rate (red arrowhead).
Fig. 32 - Tomografia computerizzata di Scipionyx samniticus, mostrante le ossa del lato sinistro. La posizione del quadrato sinistro
(freccia rossa) rispetto al destro (freccia verde) rivela che la protuberanza che deforma il quadratogiugale destro è causata dal condilo
mediale del quadrato sinistro (punta della freccia rossa).

ramus (jugal process) is still firmly sutured to the jugal,
whereas the ascending ramus (quadrate process) has lost
its medial contact with the quadrate as a consequence of
diagenesis. The ascending ramus terminates in a simple
tapering point without reaching the descending process
of the squamosal. As noted below (see Squamosal), this
simple shape is often related to a point contact, at most,
with the squamosal.

A quadratojugal similar to that of Scipionyx can be
seen in Eoraptor, Caudipteryx and Sinornithoides (Rus-
sell & Dong, 1993). At least under UV light, a similarly
shaped quadratojugal is also present in Juravenator (G6h-
lich & Chiappe, 2006: fig. 2a, contra fig. 2b). In the UV
picture, the rami are seen tapering to a pointed end, with
the jugal ramus clearly accommodated in a forked depres-
sion of the jugal. The two rami seem to be of equal length
and form a right angle. Unfortunately, the quadratojugal
is not well-exposed in the other known compsognathids,
limiting the possibility of comparison.

A knob-like, rounded relief is visible in the corner be-
tween the two rami (Figs. 23-24). At first glance, it some-
what resembles the cornual process of the jugal visible in
Alioramus (Brusatte et al., 2009), but it is not part of the
quadratojugal. Rather, as shown by CT analysis, this re-
lief is the result of the diagenetic compression of the thin
quadratojugal onto the medial condyle of the left quadrate
(Fig. 32).

Squamosal - The squamosal of Scipionyx is com-
posed of four prominent processes developed in three dif-
ferent planes: dorsal, lateral and caudal (occipital) (Fig.
33). In the dorsal plane, the squamosal is in extensive
contact with the parietal medially (parietal process) and
borders the supratemporal fenestra rostrally. On account
of some crushing, the postorbital process appears to be
bifurcated: the upper ramus belongs to the dorsal surface

of the squamosal and borders the caudolateral margin of
the supratemporal fenestra; the lower ramus belongs to
the lateral surface of the bone and forms the dorsal mar-
gin of the infratemporal fenestra. The lateral surface of
the squamosal includes also a robust descending process
(quadratojugal or paraquadrate process) which terminates
in a squared apex and seems to partially reduce the caudal
margin of the infratemporal fenestra.

Ppsq

capsq

Fig. 33 - Shaded drawing of the right squamosal of Scipionyx samniti-
cus, and its tetraradiate structure. See Appendix 1 or cover flaps for
abbreviations.

Fig. 33 - Disegno ombreggiato dello squamoso destro di Scipionyx
samniticus, con la sua struttura tetraradiata. Vedi Appendice 1 o risvolti
di copertina per le abbreviazioni.

44 CRISTIANO DAL SASSO & SIMONE MAGANUCO

The descending process is probably distorted and
might have paralleled the quadrate shaft (but see below)
and contacted the ascending process of the quadratojugal
in the uncrushed skull. A similar contact occurs in Jurave-
nator (GGhlich & Chiappe, 2006: fig. 2). Ifthe descending
process was naturally formed the way we observe it, in
Scipionyx the squamosal and quadratojugal did not con-
tact each other. This condition is seen in some basal sau-
rischians, such as Eoraptor and Herrerasaurus (Langer,
2004); in some non-tetanuran theropods, such as Di/o-
phosaurus and some abelisaurids (Rauhut, 2003); in the
advanced Maniraptora, such as the troodontids Mei (Xu &
Norell, 2004) and Sinornithoides (Russell & Dong, 1993);
in the dromaeosaurids Deinonychus (Ostrom, 1969) and
Sinornithosaurus (Xu & Wu, 2001); and in birds (Padian,
2004), Archaeopteryx included (Elzanowski, 2002). There
is a particular similarity between Scipionyx and Herrera-
saurus, as in both genera the descending process of the
squamosal terminates in a squared end which expands
rostrally within the caudal portion of the infratemporal
fenestra. Kobayashi & Barsbold (2005) pointed out that
the tip of the ventral process of the squamosal is missing
in the ornithomimosaur Garudimimus, but it is interesting
to note that in both figs. 2 and 3 of Kobayashi & Barsbold
(2005) the length and the shape of the process is the same
in right and left squamosals. Even if we take into account
erosion of the apexes on both sides, it is possible that the
descending process in Garudimimus is almost complete
and that it terminated in a squared end that did not contact
the quadratojugal. The descending process is rather short
and triangular in closely related forms such as Shenzhou-
saurus (Ji et al., 2003) and Sinornithomimus (Kobayashi
& Lii, 2003), even if in the latter the squamosal is reached
by the very long squamosal process of the quadratojugal.

In Scipionyx, irrespective of whether the squamosal
and the quadratojugal contacted or not, the quadrate would
have been positioned beneath the quadratojugal, resulting
more inclined, so that the descending process of the squa-
mosal would have been less inclined with respect to the
quadrate than it appears now. For this reason, character
48 has been coded (0) for Scipionyx in our phylogenetic
analysis.

Caudal to the base of the quadratojugal process, a deep
cotyle accommodates the head of the quadrate. Starting
from the cotyle, the caudal process of the squamosal is
developed caudoventrally, ending in a thin bony crest.

Frontal - The frontal is a large bone with a smooth
surface. It is rostrocaudally elongate and widest along the
caudal edge, where it forms a distinct postorbital process.
As for the nasal, diagenesis has caused the right frontal to
appear flattened, its lateral and dorsal sides now lying on
the same plane. As described above, the frontal articulates
rostrally with the nasal via a thin interdigitate suture. Ros-
troventrally, it articulates with the prefrontal via a linear
suture. The left frontal-prefrontal contact has slid about 2
mm ventrally and is now exposed in medial view, in the
dorsalmost area of the orbit. The left frontal, visible as a
crescent-shaped surface along the dorsal rim of the right
orbit, was wrongly identified as the right “inner (orbital)
wall of the frontal” by Dal Sasso & Signore (1998a).

A large U-shaped notch marks the mediocaudal mar-
gin of the frontal and the rostromedial margin of the pari-
etal, just medial to the contact between the two bones. It

was formerly thought that this notch was created during
the initial preparation of the specimen, when some bone
fragments were lost (Fig. 19A). Three-dimensional re-
construction of the skull (Fig. 34) revealed that this notch
would have been smaller in the living animal, indicating
that part of it has derived from the diagenetic crushing of
the hemispheric cranial vault, and that the remnant gap
represents, in fact, the still open frontoparietal fontanelle
(see Ontogenetic Assessment). The margins of the fron-
toparietal fontanelle do not appear broken. Likely, the
frontoparietal fontanelle acted as a weak point and fa-
voured separation of the frontal and parietal bones during
diagenesis.

Ventral to the notch, and up to the postorbital, the
frontoparietal suture parallels the margin of a bony ridge
having a rugose texture and a sinuous course. This ridge
forms the caudolateral end of the frontal and marks the
rostral limit of both supratemporal fossa and fenestra.
It was previously described as a ‘“transverse postorbital
ridge” (Dal Sasso & Signore, 1998a), but should be more
properly defined as a “sinusoidal ridge of the supratempo-
ral fossa” (Currie, pers. comm., 1998). A similar structure
is considered a diagnostic feature of the dromaeosaurids
(Currie, 1995; Barsbold & Osmolska, 1999; Xu & Wu,
2001; Norell & Makovicky, 2004), and was described in
troodontids (Makovicky & Norell, 2004) and in the basal
tyrannosauroid Guanlong (Xu et al., 2006). In all these
taxa, the ridge lies in a position more rostral than that in
Scipionyx, separating the skull table from the excavated
caudal tip of the frontal, that participates in the supratem-
poral fossa. Functionally, this crest might have acted as a
supplementary attachment site for the temporal muscula-
ture (Holliday, 2009). A sinuous frontoparietal suture is
also present in Compsognathus (Peyer, 2006), Sinornitho-
saurus (Xu & Wu, 2001) and in the ornithomimosaur Si-
nornithomimus; in the latter taxon it seems associated with
a low, rugose ridge (Kobayashi & Li, 2003: fig. 7B).

Fig. 34 - Cardboard model of the skull of Scipionyx samniticus, show-
ing the frontoparietal notch of the fossil reduced to a narrow space,
which is consistent with the fontanelle of a hatchling individual. See
Appendix 1 or cover flaps for abbreviations.

Fig. 34 - Modello in cartoncino del cranio di Scipionyx samniticus. Con
le ossa ricomposte, l’ampia incisura frontoparietale vista nel fossile si
riduce ad uno spazio stretto, compatibile con la fontanella di un indi-
viduo appena uscito dall’uovo. Vedi Appendice 1 o risvolti di copertina
per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 45

Scipionyx resembles most ornithomimosaurs (e.g.,
Kobayashi & Lii, 2003), troodontids (e.g., Ji ef a/., 2005) and
birds (Rauhut, 2003) in having frontals that (1) flare over the
orbits up to the frontoparietal suture, where the skull reaches
its maximum width, and (2) become domed near the caudal
part of the orbit, where the skull forms a flexure between the
flat parts of the frontals and the parietals.

In Scipionyx, the frontal forms an extensive portion
of the dorsal margin of the orbit in lateral view, as does
the frontal ramus of the postorbital. In Compsognathus
(Peyer, 2006), the frontal does not extend beyond the dor-
socaudal quarter of the orbit. A condition similar to that of
Scipionyx can be seen in Juravenator and is common in
many other unrelated taxa, such as Eoraptor, Coelophysis,
Nqwebasaurus (de Klerk et al., 2000), oviraptorosaurs,
therizinosauroids, ornithomimosaurs (e.g., Kobayashi &
Barsbold, 2005; Kobayashi & Lii, 2003) and deinonycho-
saurs (Xu & Wu, 2001; Barsbold & Osmélska, 1999). Un-
fortunately, the orbital area is not preserved well enough
in other compsognathids.

Parietal - Diagenetic crushing has caused three por-
tions of the parietal of Scipionyx to fall on the same plane
rather than being exposed in three different views as oc-
curred in life: these are an extended dorsal portion, which
is almost flat and lacks any evident sagittal crest; a cau-
dolateral (occipital) portion, which is smaller than the
dorsal portion and is slightly concave towards the medial
plane; and a ventral portion, partly visible through the up-
per half of the supratemporal fenestra.

Rostrally, the dorsal portion is still in contact with the
sinusoidal ridge of the frontal for most of its length, as
first noted by Currie (pers. comm., 1998). Towards the
postorbital, it tapers forming a triangular tip, which rep-
resents the postorbital process and overhangs the rostral
portion of the supratemporal fossa. This tip has lost its
contact with the frontal because of crushing of the skull
during diagenesis: this opened the parietal like a fan and
produced the small crack that can be seen along the dor-
sal rim of the supratemporal fenestra. As a matter of fact,
both crack and contact between the postorbital process of
the parietal and the frontal become automatically restored
in reconstructing the vault in a three dimensional model.
Caudally, the dorsal portion of the parietal is in continuity
with the lateral half of the caudolateral portion, whereas
in the medial half it has lost its contact with the rostral
margin of the supraoccipital, to which it is apparently
complementary. The loss of contact is due to crushing,
which produced also the crack visible 2 mm rostrally.

Contact with the supraoccipital is still maintained by
the caudolateral (occipital) portion of the parietal, which
Dal Sasso & Signore (1998a) attributed to the supraoc-
cipital. As mentioned, in the caudal plane this small por-
tion of the parietal forms a medially oriented concavity,
the fossa for the ligamentum nuchae, and contacts the
supraoccipital via an S-shaped suture, which delimits the
fossa and is more evident under UV light (Fig. 35). Later-
ally, the parietal contacts the squamosal through a long,
straight suture.

The ventral portion of the parietal appears as a flange
below the dorsal portion, terminating ventrally in a clean-
cut horizontal suture with the laterosphenoid. The parietal
and the laterosphenoid together form the lateral wall of
the braincase, and caudally they contact the prootic.

Fig. 35 - Close-ups of the right occipitoparietal region of Scipionyx
samniticus under ultraviolet-induced fluorescence (A) and visible light
(B), documenting the position of the fossa for the ligamentum nuchae,
and the complex shape of the suture between the supraoccipital and the
parietal. Scale bar = 1 mm. See Appendix 1 or cover flaps for abbrevia-
tions.

Fig. 35 - Particolari della regione occipitoparietale destra di Scipionyx
samniticus, in fluorescenza indotta da luce ultravioletta (A) e in luce
visible (B), che documentano la posizione della fossa per il legamento
della nuca, nonché il complesso andamento della sutura tra sopraoc-
cipitale e parietale. Scala metrica = 1 mm. Vedi Appendice 1 o risvolti
di copertina per le abbreviazioni.

In Scipionyx, the parietals lack an evident sagittal crest
and have only a weakly marked transverse nuchal crest at
the angle between the skull roof and the occipital plane.
This condition is certainly expected in an immature speci-
men, but it is noteworthy that it is not necessarily related
to the ontogenetic stage, as feebly developed parietal
crests are found in adult individuals of small-sized thero-
pods belonging to different groups that are not strictly re-
lated to each other, such as the coelophysoids, the comp-
sognathids, the ornithomimosaurs (in which the sagittal
crest is totally absent; [Makovicky et a/., 2004]) and the
oviraptorosaurs.

Braincase

The braincase elements are largely obscured by
the surrounding bones of the dermal skull roof: stapes,
sphenethmoidal and basipterygoid are entirely invisible.
However, some crushed and partly broken elements are
visible through the orbital, supratemporal and infratem-
poral fenestrae.

As for the palate described below, the lack of pub-
lished descriptions of the braincase in basal coelurosauri-
ans renders it difficult to compare this region and evaluate
its eventual taxonomic significance. This is the case, for
example, for the tympanic recesses. So, even when ex-
tending the comparison to other theropods, we had to base
the interpretation of this area on only a few specimens:
Sinraptor (Currie & Zhao, 1993a), Sinovenator (Xu et
al., 2002b), Troodon (Currie & Zhao, 1993b), Gallimimus
(Osmolska et al., 1972) and Majungasaurus (Sampson &
Witmer, 2007).

Supraoccipital - As mentioned above, the supraoccip-
ital has lost contact with the parietal medially. The right
supraoccipital begins rostrally with a transverse ridge,

46 CRISTIANO DAL SASSO & SIMONE MAGANUCO

reinforcing the sutural margin, which can be traced as a
relief also on the thin portion of the left supraoccipital
exposed just rostrally to the right one (Fig. 35). The ridge
marks the transition between the dorsal and the occipital
plane, the surface of which is marked by a rugose texture
for the attachment of the nuchal muscles.

As in the parietal, there is no trace of a sagittal crest
close to the midline in the rostralmost portion of the su-
praoccipital. The lateral margin of the supraoccipital bor-
ders on the concavity present on the caudolateral surface
of the parietal.

In the holotype of Scipionyx, the skull is firmly ar-
ticulated to the vertebral column, so that the ventralmost
portion of the supraoccipital terminates juxtaposed to the
neural arch of the atlas. Nevertheless, the ventral margin
of the supraoccipital is visible. For a short, medial tract it
could represent the dorsal arch of the foramen magnum,
whereas the remaining portion could be the exoccipital
suture. This uncertainty is due to the fact that the occipital
area is not known or not well-described in compsogna-
thids, whereas in some other basal coelurosaurs, such as
the basal tyrannosauroid Guan/ong (Xu et al., 2006), the
supraoccipital does not border the foramen magnum.

The supraoccipital of Scipionyx closely resembles that
of Allosaurus (Madsen, 1976: fig. 13) in general shape
and mode of contact with the parietal. Sinocalliopteryx is
the only other compsognathid in which the supraoccipital
is well-exposed (Ji ef a/., 2007a: fig. 2a). It bears a sagit-
tal crest and, according to Ji ef al. (2007a), it has also a
pair of fossae that possibly represent the fossae ligamenti
nuchae, as is the case in Velociraptor (Barsbold & Os-
molska, 1999). Scipionyx is similar to many basal teta-
nurans (Holtz et al., 2004) in that these fossae are present
on the parietals. A foramen, which by comparison would
represent the entrance of the right vena capitis dorsalis
(e.g., Rauhut, 2003), opens in the wall of the supraoc-
cipital of Scipionyx, lateral to the right fossae ligamenti
nuchae (Fig. 35).

Exoccipital - The exoccipitals of Scipionyx are largely
hidden by the right squamosal; only the end of the right
paraoccipital process can be seen. We are more confident
in interpreting the exposed surface as the caudal (occipi-
tal), rather than the rostral one, as it is smooth, lacking any
rugose texture, is slightly concave and is not marked by
an extensive sutural contact with the squamosal.

The exposure of the occipital surface of the paraoc-
cipital process in a skull in lateral view is certainly unu-
sual. However, as the quadrate and the articular region of
the mandible are visible in lateral view, rather than being
hidden by the quadratojugal, it is apparent that, during
diagenesis, deformation of this area of the skull occurred
with movement of the bones in a rostral direction. Thus,
the lateral end of the paraoccipital process would have
been dragged rostrally too. The ridge with rounded mar-
gins which borders this element is attributable to the ven-
tral and lateral margins of a paraoccipital process, where-
as the outline of the bone visible as a relief underneath the
quadrate matches the shape of the dorsal margin.

Prootic - A drop-shaped dimple with a rostrally di-
rected apex is visible in the dorsocaudal corner of the in-
fratemporal fenestra (Fig. 36). This dimple represents the
fenestra ovalis (the cavity at the bottom of the middle ear,

accommodating the head of the stapes). It is delimited by
two small crests, one of which — that ventral to the dimple
— is the crista interfenestralis. These structures identify
the bone which bears them as being the prootic. The pres-
ence of what would be the dorsal tympanic recess, which
opens dorsally to the fenestra ovalis, supports this inter-
pretation. The partly visible recess which opens ventral to
the crista interfenestralis might be the subotic recess.

Most of these elements were well-illustrated and de-
scribed in the troodontid Sinovenator (Xu et al., 2002b).
In Scipionyx, the dorsal tympanic recess is directly vis-
ible only for a short tract, after which its outline can be
traced underneath the squamosal as a depression (Figs.
24, 36B) running up to the supratemporal fenestra, where
it surfaces again in the caudoventral corner. This inter-
pretation is consistent with prootic recesses with similar
topology, i.e., adjacent to the parietal and laterosphenoid,
observed in coelophysoids (Raath, 1985), ornithomim-
ids (Makovicky & Norell, 1998), Velociraptor (Barsbold
& Osmolska, 1999) and, possibly, Archaeopteryx (El-
zanowski, 2002). Thus, the prootic of Scipionyx extends
dorsally and is visible through the supratemporal fenestra,
where it forms a vertical suture with the laterosphenoid
and the parietal.

Basioccipital-Basisphenoid - Only a small portion of
this complex can be seen emerging just ventral to the right
paraoccipital process. Based on its size and shape, we in-
terpret this as the right basal tuber.

Basisphenoid-Parasphenoid - The most visible
structure of this bone complex is the cultriform process
of the parasphenoid. It seems to be complete and ori-
ented rostrodorsally, crossing the dorsal half of the orbit
obliquely. The cultriform process is a straight, slender,
rod-like bone that in Scipionyx, as in most small thero-
pods (e.g., Syntarsus, Dromaeosaurus), probably point-
ed rostrally. Its rostrodorsal orientation in the fossil is
likely an artefact of preservation, as it occurs only in
non-coelurosaurian taxa which attained medium to large
body size (e.g., Majungasaurus and Allosaurus). Moreo-
ver, in those taxa the cultriform process has a different,
plate-like shape (Rauhut, 2003). The simultaneous ex-
position of both right lateral surface and median ventral
sulcus indicates that in Scipionyx the cultriform process
rotated on its long axis (Currie, pers. obs., 1999). The
flat lateral wall and the presence of a ventral sulcus sug-
gest also that the cultriform process of Scipionyx had the
shape of an inverted V in cross section.

Ostrom (1978) recognised the cultriform process in
Compsognathus as a double structure with an inverted
V-shaped section, after having observed rostral laminae
unquestionably continuous with other ventral elements of
the braincase. More recently, Xu & Wu (2001) described
a triangular cross section in the cultriform process of Si
nornithosaurus.

As the base of the cultriform process is not visible, it
is Impossible to ascertain its degree of pneumatisation.
However, the exposed portion suggests that the cultri-
form process of Scipionyx does not possess a pneumatic
bulla as expanded as in troodontids (e.g., Currie & Zhao
1993b).

The cultriform process and the body of the basisphe-
noid-parasphenoid are two braincase elements that are

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 47

feo

cif

Fig. 36 - Some bones of the braincase of Scipionyx samniticus can be seen through the right supratemporal fenestra (A), including the
prootic (B). Scale bar = 1 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 36 - Attraverso la finestra sopratemporale di Scipionyx samniticus sono visibili alcune ossa della scatola cranica (A), tra cui il
prootico (B). Scala metrica = 1 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

usually fused in the archosaurs. Close examination of the
area caudal to the postorbital-jugal contact and just ven-
tral to the quadratic ala of the pterygoid reveals a mor-
phological continuity between them also in Scipionyx.
Moreover, the basisphenoid-parasphenoid complex faces
more extensively dorsally to the quadrate ala of the ptery-
goid, within the infratemporal fenestra.

Laterosphenoid - Within the supratemporal fenestra,
a partially covered bone can be seen, dorsally separated
from the parietal by a weakly incised horizontal suture.
Based on its close relationship with the right parietal, it
can be interpreted as the right laterosphenoid. Contra Dal
Sasso & Signore (1998a: fig. 4), the bone emerging in
the dorsal third of the infratemporal fenestra is not the
laterosphenoid, but part of the prootic and basisphenoid-
parasphenoid complex, as described above.

Orbitosphenoid - Microscopic examination of the el-
ement previously identified as the orbitosphenoid by Dal
Sasso & Signore (1998a) revealed that it has a laminar
structure and texture analogous to scleral plates. That ele-
ment can now be ascribed, therefore, to the right scleral
ring, to which it has also a compatible position. Although
it cannot be proven, it is possible that a thin portion of the
real orbitosphenoid of Scipionyx is visible in the same ar-
ea, in the form of a splinter of bone previously considered
as the rostromedial (inner) wall of the postorbital (‘“ipo”
in Dal Sasso & Signore, 1998: fig. 4).

Palatoquadrate complex

Vomer - In theropods, the vomers — thin rods of bone
located astride the sagittal midline of the palate — are gen-
erally indistinguishably fused rostrally (see Ontogenetic
Assessment). However, in the holotype of Scipionyx they
are completely separated and well-exposed (Fig. 37). The
right vomer seems to be in contact with the left palatine via
a laminar portion of the pterygopalatine process, but it is
only superimposed on it. In fact, the right vomer has rotat-
ed clockwise up during diagenesis to expose its rostral end
(premaxillary process) in-between the m4 teeth, where it
overlaps the left vomer (Fig. 30). Because of this rotation,
the pterygopalatine end now points dorsally, but remains
still aligned with the main body of the bone, showing that
the vomer of Scipionyx is back straight. The left vomer,
which is not so well-preserved, has lost contact with both
palatine and pterygoid: facing rostrally between m2 and
m4, it runs caudally in a straight line under the crown of
md, re-appearing first between m4 and mS, where it passes
below and under the right vomer, and then in the caudo-
ventral corner of the antorbital fenestra, and finally slip-
ping underneath the left palatine (Figs. 25C-D, 30).

Palatine - Both palatines are exposed, the right one
within the rostral portion of the orbit, and the left one
within the caudal half of the antorbital fenestra. There-
fore, the right palatine has shifted dorsocaudally with re-
spect to the position it had in life, together with the whole

48 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 37 - Shaded drawing of the palatines and vomers of Scipionyx samniticus. See Appendix 1 or cover flaps for abbreviations.
Fig. 37 - Disegno ombreggiato dei palatini e vomeri di Scipionyx samniticus. Vedi Appendice 1 o risvolti di copertina per le abbre-

viazioni.

pterygoid complex, whereas the left palatine has only
shifted dorsally. The main body of the palatine (intercho-
anal bar) terminates dorsally in a margin that is unlike
the large, lobate structure of A//osaurus (Madsen, 1976).
The right lacrimal obscures the maxillary process and the
vomeral process of the right palatine. Nevertheless, the
shifting of the right palatine has revealed the pterygoid
process and the jugal process. Instead, the parts of the left
palatine that are visible are the thin maxillary process and
the vomeral process, which mark the caudal margin of
the choana. Thus, the combined observations on the two
palatines demonstrate that these bones are tetraradiate in
Scipionyx (Fig. 37).

A tetraradiate palatine (i.e., possessing also a jugal
process) is partially exposed also in Compsognathus
(Peyer, 2006: fig. 4A). Apart from this datum, the other
known compsognathids do not show any other detail use-
ful in reconstructing the shape of their palatine.

In Scipionyx, the pterygoid-palatine contact is very limit-
ed on account of the presence ofa subsidiary palatal fenestra.
Seemingly, the left palatine contacts the right vomer, which
pivoting on the apex of the vomeral process of the palatine,
has rotated clockwise. Because of this rotation, part of the
vomeropalatine suture cannot be distinguished. Differently
to A/losaurus (Ji et al., 2003; Norell & Makovicky, 2004)
and Daspletosaurus (Currie, 2003), the vomeral process is
longer than the maxillary process in Scipionyx, as is the case
in Majungasaurus (Sampson & Witmer, 2007), Sinraptor,
Shenzhosaurus, Deinonychus, Velociraptor and, possibly,
Juravenator (GOhlich & Chiappe, 2006: fig. 2a).

Within the notch delimited by the maxillary process
and the jugal process, some plates of the ?right scleral ring
are found. The palatine and the ectopterygoid are not com-
pletely separated by the pterygoid, having a point contact
at most. This contact is more extensive in Velociraptor
than in Scipionyx, and even more in ornithomimids and

oviraptorosaurs, in which the two bones contact through a
long oblique suture (Barsbold & Osmòdlska, 1999).

Pterygoid - The pterygoid is the longest cranial bone of
Scipionyx, extending rostrocaudally for about two-thirds of
the whole cranial length (Figs. 25C-D, 38). Both right and
left pterygoids have markedly shifted dorsally above the
plane of the palate. Now their palatal rami, still separated by
the interpterygoid vacuity, cross the orbit horizontally, con-
verging rostrally and intersecting beneath the lacrimal: they
re-emerge in the dorsal portion of the antorbital fenestra,
diverging up to their rostral ends (vomeral processes).

In both pterygoids, the rod-like palatal ramus is well-
preserved, whereas the laminar portion formed by the lat-
eral processes directed towards the ectopterygoids and the
palatines is preserved and partially visible only in the right
pterygoid. It has a rostrolaterally directed horizontal process
(caudal palatine process) which, contacting the medial wall
of the palatine, delimits two fenestrae: the subsidiary palatal
finestra rostrally and the suborbital fenestra, which is mark-
edly larger, caudally. As mentioned, the subtemporal fenestra
is delimited rostrally by another process of the pteryogoid
(ectopterygoid process), which rises from the caudalmost
portion of the palatal ramus and forms an arched bar ventro-
laterally directed and contacting the ectopterygoid. In the un-
crushed skull, the ventralmost end of this bar, now covered
by the jugal, would have projected down from the roof of
the mouth, ventral to the jugal, like in many theropods (e.g.,
Weishampel et a/., 2004). At the same level, the ventromedi-
al surface of the arch forming the left ectopterygoid process
can be seen between the two palatal rami of the pterygoids.
Caudally, the horizontal rod of the right palatal ramus termi-
nates in a pointed caudomedial process (posteromedial proc-
ess sensu Madsen & Welles, 2000: pl.5) which, because of
diagenetic compression, is visible as a relief in lateral view,
slightly superimposed onto the lateral surface of the jugal.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 49

Fig. 38 - Shaded drawing of the right pterygoid and ectopterygoid of Scipionyx samniticus. See Appendix 1 or cover flaps for ab-

breviations.

Fig. 38 - Disegno ombreggiato dello pterigoide ed ectopterigoide destri di Scipionyx samniticus. Vedi Appendice 1 o risvolti di coper-

tina per le abbreviazioni.

An abrupt, right-angled change of direction in the cau-
domedial process of the pterygoid forms a vertical ramus.
This vertical ramus is reinforced by a ridge which expands
in the postorbital region to form a broad, bony lamina called
the quadrate ala, previously interpreted as the basisphenoid
(“bs” in Dal Sasso & Signore, 1998a). This ala occupies
more than one third of the infratemporal fenestra and con-
tacts the quadrate with a broad, extensive suture.

In the middle of the infratemporal fenestra, the quadrate
ala protrudes up to reach a plane more raised laterally than
the postorbital-jugal bar, exposing all its thickness and a ros-
tral margin which matches perfectly the outline of the caudal
margin of the epipterygoid. This margin, therefore, repre-
sents the contact margin between the two bones (Fig. 25C).

Epipterygoid - In lateral view, the epipterygoid of
Scipionyx has roughly the outline of an isosceles trian-
gle, with the acute apex slightly bent rostrally (Fig. 39).
It is fossilised with the caudal margin leaning against the
rostromedial (inner) wall of the postorbital, but in the
undisturbed skull, as mentioned, it linked the quadrate
ala of the pterygoid to the laterosphenoid. Ventrally, the
epipterygoid forks into a rostral process, which originally
was in continuity with the reinforced rostral ridge of the
quadrate ala of the pterygoid, and a caudal process of sim-
ilar shape and size. In between these processes, the outer
surface of the epipterygoid bears a shallow concavity.

This bone was not identified by Dal Sasso & Signore
(1998a: fig. 4), who interpreted the dorsal half of the bone
as part of the postorbital (“ipo”), leaving the ventral half
as an indeterminate bone fragment. The epipterygoid of
Scipionyx is similar in shape to that of A//osaurus (Mad-
sen, 1976) and Daspletosaurus (Currie, 2003).

Ectopterygoid - Because of the compression of the
skull, the right ectopterygoid of Scipionyx exposes mainly
its dorsal surface and emerges in the orbit, just below the
right pterygoid. They remain in contact through a large,
T-shaped pterygoid process (Fig. 38). The same shape and
mode of contact can be seen in Juravenator (GGhlich &
Chiappe, 2006: fig. 2a).

Rostrally, the ectopterygoid borders the caudal mar-
gin of the suborbital fenestra and has a point contact with
the pterygoid process of the palatine. The contact with the
rod of the palatal ramus of the pterygoid is markedly ex-
tensive, and more so than in Sinraptor (Currie & Zhao,

Fig. 39 - Close-up of the right postorbital region of Scipionyx samniti-
cus, showing the diagenesis-related protrusion of some elements of the
braincase and of the palate above the level of the postorbital bar. Scale
bar = 1 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 39 - Particolare della regione postorbitale destra di Scipionyx sam-
niticus. Si noti la protrusione, dovuta alla diagenesi, di alcuni elementi
della scatola cranica e del palato al di sopra del livello della barra
postorbitale. Scala metrica = 1 mm. Vedi Appendice 1 o risvolti di co-
pertina per le abbreviazioni.

50 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 40 - Palatal bones of Scipionyx samniticus under ultraviolet (A)
and visible light (B). Note the extensive contact between the pterygoid
and the ectopterygoid. Scale bar = 2 mm. See Appendix 1 or cover flaps
for abbreviations.

Fig. 40 - Ossa del palato di Scipionyx samniticus in luce ultravioletta
(A) e visibile (B). Si noti il contatto prolungato tra lo pterigoide e
l’ectopterigoide. Scala metrica = 2 mm. Vedi Appendice 1 o risvolti di
copertina per le abbreviazioni.

1993a). As in Sinraptor, another fenestration is present in
Scipionyx, between the ectopterygoid process of the ptery-
goid and the pterygoid process of the ectopterygoid. This
is unlike Dromaeosaurus, in which a long, continuous su-
ture unites the pterygoid caudally with the ectopterygoid.
Besides part of its bulky portion, most of its lateral por-
tion forming the jugal process and its caudoventral corner,
which contacted the ventral bar of the pterygoid project-
ing below the roof of the mouth, are hidden by the jugal.
Under UV (Fig. 40A) and properly oriented visible light
(Fig. 40B), a shallow circular depression can be seen at
the centre of the main body of the ectopterygoid. This de-
pression is consistent with the position of the pneumatic
opening, or ventral fossa, which opens ventrally in many
theropods (e.g. Currie, 2003: fig. 12).

Quadrate - In lateral view, the quadrate of Scipionyx
appears as a relatively tall, subvertical pillar with rounded
ends (Fig. 41). Actually, it possesses also a remarkable
expansion, the pterygoid ala, directed rostromedially and
partially covered by the descending process of the squa-
mosal and the ascending process of the quadratojugal. As
mentioned, the pterygoid ala consents the pterygoid and
the quadrate to be in perfect anatomical continuity, form-
ing the medial wall of the adductor chamber.

The outline of the ventral margin of the pterygoid ala
is well-delineated only in proximity of the lateral condyle;
rostroventrally, it is hardly distinguishable from the lateral
surface of the caudal portion of the mandible because of
the similarities in surface texture and the strong diagenet-
ic compression which almost fused the laminar portion of
the quadrate onto the surangular. For this reason, part of
the pterygoid ala was misinterpreted as the caudal portion
of the pterygoid by Dal Sasso & Signore (1998a: fig.4).

The probably single quadrate head is exposed in lateral
view within the laterally open quadrate cotyle of the squa-
mosal, slightly caudal to the vertical projection of the quad-
rate lateral condyle. In fact, during diagenesis the quadrate
pivoted on the quadrate cotyle of the squamosal, loosing
the connection with the quadratojugal and rotating a little
caudally. As a consequence, the quadrate is now faintly
inclined rostrally, whereas in vivo it was slightly more in-
clined, based on the position of the quadratojugal which is
still articulated to the jugal. According to Rauhut (2003), a
rostral inclination of the quadrate occurs in various degrees
in spinosaurids, ornithomimosaurs and Archaeopteryx.

The caudoventral corner of the lateral condyle presents
a rugose, drop-shaped surface which represents the ar-

ticular surface for the quadratojugal, confirming the post
mortem loss of articulation between the two bones. A little
concavity marks the paraquadrate foramen just dorsal to
that articular surface; this permitted the passage of neuro-
vasculature between the occiput and the adductor chamber.
No remains of the left quadrate are directly visible in
Scipionyx. However, as mentioned, the low dome visible
under the right quadratojugal (Figs. 23, 42) represents the
medial condyle of the left quadrate. This is shown by the
deepest parasagittal slices of the CT scan (Fig. 32).

Mandible

The bones of the lower jaw of Scipionyx formed two
slender, straight and laterally compressed rami united at
the rostral ends by a short, weak symphysis. The tooth
row of the mandible extends from the symphysis to the
level of the caudal border of the antorbital fenestra and,
notably, further caudally than the upper tooth row (Figs.
23-25; see also Remarks to the Emended Diagnosis).

Contrary to what appeared to be logical on first exami-
nation of the fossil (Dal Sasso & Signore, 1998a), the bones
of the mandible protruding most to the right side actually
belong not to the right hemimandible but rather to the left.
Looking at the mandible in a very oblique right ventrolat-
eral view (Fig. 42), the left mandibular ramus can be seen
to stick out from the plane of the slab more than the right

Fig. 41 - Shaded drawing of the right quadrate of Scipionyx samniticus.
See Appendix 1 or cover flaps for abbreviations.

Fig. 41 - Disegno ombreggiato del quadrato destro di Scipionyx sam-
niticus. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY A Sil

one, from only few millimetres caudal to the symphysis
backwards. Thus, the left mandibular ramus shifted ventral
to the right ramus during diagenesis, and then was crushed
in its dorsal half under the right ramus itself; as a conse-
quence, it remained raised in the ventral half. In the end, the
bones composing that portion (1.e., left dentary, left splenial
and left prearticular) lie on a plane about 0.3 mm higher
than that of the counterlateral elements.

Based on what can be seen in both rami, almost all the
elements of the lower jaw can be described and recon-
structed, at least in part. The only exception is represented
by the coronoids, which in Scipionyx are covered by the
bones of the skull and of the right hemimandible.

Mandibular openings - In lateral view, the mandible of
Scipionyx lacks any large opening. Contrary to Dal Sasso
& Signore (1998a: fig. 4, “emf”) and according to Peyer
(2006), there is no external mandibular fenestra: actually,
the large depression below the right angular originated from
the dorsal displacement of that bone. This depression has
brought to light the lateral surface of the right prearticular.
There is no external mandibular fenestra also in the lower
Jaw of the compsognathids Compsognathus (Peyer, 2006),
Huaxiagnathus (Hwang et al., 2004), Sinosauropteryx (Cur-
rie & Chen, 2001) and Juravenator (G6hlich & Chiappe,
2006), or in Archaeopteryx (Elzanowski, 2002). According
to Peyer (2006), the absence of a mandibular fenestra is an
autapomorphy of compsognathids. Interestingly, this open-
ing is reported to be absent or extremely reduced also in the
basal tyrannosauroid Di/ong (Xu et al., 2004).

Evidence for an adductor fossa facing medially in the
caudal third of the lower jaw (Fig. 42), and bordered dor-
sally by the surangular and ventrally by the prearticular, is
given by the reinforced margin of the right surangular and
the convex rim of the left prearticular.

The Meckelian fossa is not exposed because the sple-
nial, which covers it, is preserved in situ; however, obvious

evidence for the presence of the fossa are the Meckelian
groove of the dentary and the mylohyoid notch (Meckel
notch) of the splenial. As for the latter, we tend to interpret
it as a notch rather than as a foramen because it is located
close to the ventral margin of the splenial, along its suture
with the dentary (see Splenial).

The presence of an internal mandibular fenestra can-
not be ascertained because the right dentary obscures the
dorsal half of the contact between the left prearticular and
the left splenial.

Dentary - The dentary of Scipionyx is quite elongate
and tapers gradually rostrally, with the ventral and the dor-
sal edge becoming subparallel. It appears also quite slen-
der, but less so than in Compsognathus (Ostrom, 1978),
Huaxiagnathus (Hwang et al., 2004), Sinosauropteryx
(Currie & Chen, 2001; Ji et al., 2007b) and Juravenator
(Gohlich & Chiappe, 2006), because the snout of Scipio-
nyx is proportionally less elongate. The bone is almost
fully toothed, with a tooth row of 10 teeth.

Rostrally, the right and left dentaries articulate
through a short subvertical symphysis, which in life
would have been quite weak, as in many theropods, al-
lowing considerable kinesis (Britt, 1991; and references
therein). Only the rostral suture can be seen of the sym-
physis, because the two rami are preserved still in articu-
lation and hide the articular surface. The rostral end of
the left dentary does not lie in medial view but, rather,
is flattened and deformed for a short tract into a flap ly-
ing on the remnants of the left dentary, maintaining its
contact with the rostral end of the right dentary. It is un-
likely that the symphysis was so firmly sutured to pro-
mote distorsion of the right dentary instead of separation
of the two rami (see also Skeletal Taphonomy). Thus,
the preserved contact between the two rami suggests that
the symphyseal region was more U- than V-shaped when
seen in occlusal view.

Fig. 42 - A grazing view of the mandibular bones of Scipionyx samniticus reveals the uplift of the elements from the left side along their
ventral margin, with respect to the same margin of the right hemimandible (follow the two margins from the symphysis, in a caudal
direction). See Appendix 1 or cover flaps for abbreviations.
Fig. 42 - Una vista radente delle ossa mandibolari di Scipionyx samniticus mostra il sollevamento degli elementi del lato sinistro lungo
il margine ventrale, rispetto allo stesso margine dell’emimandibola destra (seguire i due margini in direzione caudale, a partire dalla
sinfisi). Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

S2 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Caudally, the right dentary connects with the surangu-
lar through a zigzagging suture that, after having formed
a dorsal process (intramandibular process sensu Currie &
Zhao, 1993a), runs obliquely in a caudoventral direction,
suturing with the angular in its caudal third. Contrary to
many theropods, in which the caudal extension of the dor-
sal margin of the dentary never goes beyond the level of
the caudalmost maxillary tooth (Rauhut, 2003), in Scipio-
nyx the dorsal margin of the dentary reaches the level
of the rostral portion of the orbit. Other exceptions are
Juravenator, which seems to be similar to Scipionyx, and
Ornitholestes, in which the dorsal margin of the dentary
is even more extended caudally and reaches the vertical
projection of the orbital centre.

In lateral view, the right dentary of Scipionyx bears two
rows of neurovascular foramina, not accommodated along
any deep groove: a well-defined row of foramina extends
beneath and parallel to the dorsal margin of the dentary; a
second, less marked row of foramina extends parallel and
Just dorsal to the ventral margin of the dentary. In Comp-
sognathus (Ostrom, 1978), Huaxiagnathus (Hwang et al.,
2004: fig. 2A) and, probably, Juravenator (G6hlich & Chi-
appe, 2006: fig. 2a) the foramina are arranged in two paral-
lel rows. Two rows, with the upper one sitting in a groove,
are considered to be diagnostic of Dromaeosauridae by
some authors (Currie, 1995; Barsbold & Osmolska, 1999;
Xu et al., 2001). One row of foramina, emerging within a
deep groove, is instead present in the gracile dentary of the
basal tyrannosauroid Coelurus (Carpenter et al., 2005b).
Apart from the two rows of foramina, the external surface
of the dentary of Scipionyx is almost smooth, with the ex-
ception of the thin furrows and pits observed in the bones
of many immature individuals, especially embryos and
post-hatchings (see Ontogenetic Assessment).

The medial surface of the left dentary is exposed for
all its length. In the rostral end, just caudal to the symphy-
sis, it is markedly wavy, as in A/losaurus (Madsen, 1976).
From this wavy area, a groove (the Meckelian sulcus)
opens and directs caudally, disappearing for a short tract
and then reappearing in the splenial.

Supradentary-coronoid - Parallel to the dorsal mar-
gin of the right dentary, a thin bony lamina appears as the
background of the last four teeth of the mandible. It is
hidden ventrally by the right dentary, and caudally by the
left maxilla. Its exposed portion has the shape of a ros-
trocaudally elongate rectangle. During the present study,
this bony lamina was initially interpreted as the lateral
(inner) surface of the right supradentary, or the medial
(outer) surface of the left supradentary. Although those
interpretations cannot be ruled out, based on its position
and shape we prefer to consider it as the dorsal portion of
the right splenial (see below).

Holtz et al. (2004) reported that in some tetanurans,
such as some tyrannosaurids, allosauroids and dromaeo-
saurids, the coronoid and the supradentary form a sin-
gle, continuous bone. Other tetanurans have a small, flat,
triangular coronoid, forming the rostrodorsal margin of
the adductor fossa, and a narrow, band-like supradentary
medially overlying the interdental plates along the dorsal
portion of the dentary. The bony lamina visible in Scipio-
nyx does not match coronoids, neither in shape nor size,
as it is definitely too large and rectangular. If it was a su-
pradentary, or a single continuous supradentary-coronoid,

given its length and position it would be either largely
incomplete rostrally or the only cranial bone considerably
displaced caudally during fossilisation.

Splenial - The size and position of the thin bony lam-
ina mentioned above, emerging dorsal to the caudal half
of the right dentary, suggest that it may well represent
the dorsal portion of a splenial, as noted also by Audi-
tore (pers. comm., 2010) in making the reconstruction of
the skull (see Cranial Reconstruction), and according to
the interpretive drawing of the medial side of the lower
jaw in Compsognathus (Peyer, 2006: fig. 5B). The bony
lamina probably represents the disarticulated and dorsally
displaced right splenial, rather than the dorsal half of a
broken left splenial, in that the ventral portion of the left
splenial emerges ventral to the ventral edge of the right
dentary and it is still in articulation with the left dentary,
without bearing any trace of fracture.

The visible portion of the left splenial has the shape
of a low triangle, with the base elongate rostrocaudally
and tapering rostrally. The dorsal portion of the putative
right splenial terminates rostrally more or less at the same
vertical level where the left splenial reaches the horizontal
level ofthe Meckelian groove, marking the point in which
the splenial would have probably formed a long, thin ros-
tral extremity, with subparallel margins, composing most
of the medial wall of the Meckelian canal.

Reconstructing both hemimandibles, the splenial
does not emerge in lateral view along the ventral side
of the jaw, contrary to a previous hypothesis (Dal Sasso
& Signore, 1998a) that regarded the splenial as the right
one, emerging ventrally like in dromaeosaurs (e.g., Ma-
kovicky & Norell, 2004). Rostrally to the fracture in the
slab which crosses the whole cranium, astride the suture
with the dentary, the ventral margin of the left splenial of
Scipionyx is marked by a concavity corresponding to the
mylohyoid notch/foramen (or Meckel’s notch/foramen).
Because of the imperfect preservation of the bone in that
point, it is not possible to establish with certainty whether
it is a notch opened towards the dentary, as it seems to
be, or a foramen completely enclosed in the splenial, bor-
dered ventrally by a thin splinter of bone lost during the
formation of the mentioned fracture. A ventrally facing
mylohyoid notch is reported to be present in Compso-
gnathus (Peyer, 2006) and in some tetanurans such as A/-
losaurus and Monolophosaurus (Rauhut, 2003), whereas
in the compsognathid Sinocalliopteryx (Ji et al., 2007) and
in most theropods (Rauhut, 2003) the foramen is reported
to be completely enclosed in the splenial. Caudally, the
left splenial contacts the left prearticular via a linear, ob-
lique contact, the former bone slipping under the latter.

Surangular - The surangular forms the entire dorsola-
teral margin of the mandible caudal to the tooth row. The
right surangular of Scipionyx is preserved in place, con-
nected to the right dentary via a zigzagging suture. The
dorsal margin of the right surangular, delimiting the lower
Jaw dorsally, is in perfect continuity with that of the den-
tary. From the contact with the dentary, it disappears for a
short tract under the right jugal, then it reappears for $5mm
within the orbit, superimposed on the arched, ventrolater-
ally directed bar of the right pterygoid, and terminates in
the ventral third of the infratemporal fenestra, covering
the pterygoid ala of the quadrate. This last tract was pre-

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 53

viously ascribed to the pterygoid and the basipterygoid
(Dal Sasso & Signore, 1998a: fig. 4). The ventral portion
of the right surangular is hidden by the angular, which has
moved dorsally.

Although exposed in lateral view, the right surangular
shows a morphology typical of its medial side. In fact, the
thickened dorsal rim of the surangular, inflected medially
and building the dorsal edge ofthe adductor fossa, has been
highlighted as a relief by the diagenetic crushing. A thin
splinter of the left surangular is visible ventral to the left
prearticular at the caudal end of the mandible (Fig. 42).

Angular - The bone previously interpreted as the right
surangular (“sa” in Dal Sasso & Signore, 1998a) actually
includes also the right angular. These two bones appear
like one element because the latter has moved dorsally,
losing its sutural contact with the dentary, and has over-
lapped the ventral portion of the lateral surface of the right
surangular. The identification of the bone as the angular
is supported by the fact that its ventral rim neither termi-
nates abruptly nor tapers as a contact margin; rather, it is
very smooth (Fig. 42), as expected in a marginal bone, as
is the case of the ventral margin of the dentary. The angu-
lar appears to taper caudally but it is partially hidden by
the jugal-quadratojugal complex, so in this area its shape
and mode of contact with the adjacent bones cannot be
observed.

Prearticular - Only a portion of the lateral surface of
the right prearticular can be seen in Scipionyx, i.e., the
surface which was in contact with the right angular before
the latter moved dorsally. Further support to this interpre-
tation is given by the alignment existing between the ven-
tral margin of the right prearticular and that of the dentary.
The fact that the surface of the right prearticular lies in
a depressed plane led to misinterpret the depression as a
right external mandibular fenestra (‘“emfe” and “ipa?” in
Dal Sasso & Signore, 1998a: fig. 4).

The left prearticular is the most exposed among the
bones of the left hemimandible. As it has undergone se-
vere crushing under the cheek and the right hemimand-
ible, it is broken into three pieces, previously interpreted
as distinet elements (“an”, “pa”, and “sp” in Dal Sasso
& Signore, 1998a: fig. 4). New examination of the al-
leged sutures between these three pieces in the light of
the continuity of their surfaces, which compose a regular
dorsoventrally convex margin, and based on the shape of
the prearticular in most theropods, suggests that the three
pieces belong, in fact, to a single bone, the left prearticu-
lar. Such continuity can be clearly seen in medioventral
view (Fig. 42). This is confirmed by examination under
UV light, which reveals that the splenial-prearticular gap
is more open than the space between the three pieces and
that it is brighter in colour on account of the gap being
filled by a brightly fluorescing fine matrix.

The prearticular in Scipionyx can be reconstructed as
a rostrocaudally slender, gently arched and dorsoventrally
convex bone. As mentioned, the arched bar of the preart-
icular defines the ventral border of the adductor fossa.
Rostrally, the prearticular contacts the splenial in a long,
oblique suture. Caudally, the contact with the articular is
not easily visible in lateral view, as the latter bone has
rotated during diagenesis, exposing its dorsal surface (see
below). The mode of contact between the prearticular and

the surangular cannot be seen in Scipionyx; however, a
thin caudal extension of the dentary can be seen contact-
ing the prearticular in medioventral view.

Articular - Only one articular, preserved in dorsal
view, is exposed in Scipionyx (Fig. 43). We regard this as
the left articular, based on the facts that it lies on the same
plane of the caudal end of the left prearticular, both retro-
articular process and medial glenoid are exposed with the
latter located rostroventral to the former (i.e., rostromedi-
ally in the undisturbed skull) and the dorsalmost margin
of the former is partly covered by the right basal tuber, a
bone which lies close to the medial sagittal plane.

The retroarticular process, known as the attachment
site of the M. depressor mandibulae, appears shallow,
rounded and as long rostrocaudally as the glenoid. Being
preserved in dorsal view, it is difficult to compare its shape
with that of the caudally tapering retroarticular process
of Sinosauropteryx (Currie & Chen, 2001) and Compso-
gnathus (Peyer, 2006), albeit in Scipionyx it seems to be
shorter rostrocaudally. A reduced retroarticular process is
instead present in basal tyrannosauroids (e.g., Xu et al.,
2004; Xu et al., 2006; Carpenter e? al., 2005b), ornitho-
mimosaurs (e.g., Kobayashi & Barsbold, 2005) and dro-
maeosaurids (e.g., Xu & Wu, 2001; Barsbold & Osmdl-
ska, 1999).

In Scipionyx, the medial glenoid of the mandible is
rounded, has a concave surface, and is well-bordered me-
dially by a thick margin that is exposed in dorsomedial
view rather than in a full dorsal view, marking a change
of plane that occurred during diagenesis. The change of
plane is completed at the level of the left prearticular,
which reaches the ventral margin of the articular and is
exposed in lateral view. The contact with the surangular,
the bone which forms the lateral glenoid, is not visible
because it is covered by the right quadrate.

Because in theropods the articular is usually bulky and
well-developed in all directions, we hypothesise that the
bump in the caudoventral corner of the infratemporal fe-
nestra, covered by the pterygoid ala of the quadrate and
by the surangular, might reflect the presence underneath
of the disarticulated right articular (Fig. 42). Unfortunate-
ly, the CT scan analysis performed on the specimen did
not help to clarify this detail.

Fig. 43 - Close-up of the caudal mandibular region of Scipionyx sam-
niticus. Ultraviolet-induced fluorescence (A) highlights the contour of
the right articular (B), which is fossilised in dorsal aspect. Scale bar = 1
mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 43 - Scipionyx samniticus, particolare della regione mandibolare
caudale. La fluorescenza indotta da luce ultravioletta (A) evidenzia il
contorno dell’articolare destro (B), che è fossilizzato in norma dorsale.
Scala metrica = 1 mm. Vedi Appendice 1 o risvolti di copertina per le
abbreviazioni.

54 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Dentition

The dentition of Scipionyx consists of 5 premaxillary,
7 maxillary and 10 dentary teeth (Fig. 44). The teeth are
moderate in number (see also Ontogenetic Assessment),
pointed and feebly recurved, without constriction at the
base of the crown. According to Peyer (2006), the tooth
crowns of Compsognathus and all other compsognathids
are unique among theropods in having a backward “kink”
at 2/3 oftheir height, in being straight along their mid-parts
and being expanded at their base. The teeth of Scipionyx,
except for the largest premaxillary and the rostral dentary
ones, which have a similar but only incipient bend, are
instead feebly, but regularly recurved. Therefore, in the
light of the immaturity of the specimen (see Ontogenetic
Assessment), the condition of Scipionyx does not support
the statement of Makovicky, reported by Peyer (2006),
that “this particular tooth shape could be common to juve-
nile theropods in general, rather than to a unique character
of the compsognathids”.

The right and left tooth rows are highly symmetrical in
size, shape and position of the crowns. Whereas the first
parameter is linked in all likelihood to the very early on-
togenetic stage of the individual (i.e., first tooth replace-
ment not yet started - see Ontogenetic Assessment), the
different shape of the tooth crowns related to their differ-
ent position along the jaws indicates a certain degree of
heterodonty (see below for more details).

The upper tooth row is completely antorbital in posi-
tion, ending rostrally to the vertical strut of the lacrimal
as in the vast majority of tetanuran theropods, including
toothed birds (Rauhut, 2003). As the right and left max-
illae are well-exposed, respectively, also in ventrolater-
al and ventromedial view, we are certain that there are
no empty alveoli caudal to the left and right m7. Thus,
uniquely, in Scipionyx the caudal extension of the lower
tooth row is greater than that of the upper one (Figs. 23-
24, 25A-B, 25D, 44). This potentially diagnostic charac-
ter is discussed in Remarks to the Emended Diagnosis.
All other compsognathids, and known theropods as well,
show the usual condition of a dentary tooth row that is
shorter than the maxillary tooth row, independent of their
ontogenetic stage.

Premaxillary teeth - Scipionyx has 5 premaxillary
teeth. This tooth count is confirmed by the integrity of the
ventral edge of the right premaxilla, which clearly shows
the lateral margins of 5 alveoli, and by the fact that the left
premaxilla has moved ventrally, so that the rostralmost
tooth visible in the mouth of Scipionyx cannot belong to
the left tooth row.

The majority of theropods have 4 premaxillary teeth,
including the French Compsognathus (Peyer, 2006) and
Sinocalliopteryx (Ji et al., 2007a); this count is usually re-
garded as the primitive state in theropods (Rauhut, 2003).
The premaxillary tooth count is, therefore, comparatively
high in Scipionyx. One of the three specimens of Sino-
sauropteryx published so far possibly has 5 premaxillary
teeth but, given that few theropods have more than 4 teeth
in the premaxilla, Currie & Chen (2001) concluded that it
is most parsimonious to interpret the fifth tooth as a max-
illary tooth morphologically similar to the premaxillary
ones. The condition of Scipionyx suggests that 5 premax-
illary teeth could indeed have been present in some speci-

mens of Sinosauropteryx, indicating a certain intraspecific
variability for this character. Five premaxillary teeth are
reported also in A//osaurus (Madsen, 1976) and Neovena-
tor (Hutt ef al., 1996) among tetanurans, whereas spino-
saurids have 6-7 premaxillary teeth (e.g., Dal Sasso et al.,
2005). Interestingly, the German Compsognathus is re-
ported to have only 3 premaxillary teeth (Ostrom, 1978).

The right premaxillary teeth of Scipionyx are all intact,
except for the apex of pm1, which has detached and leans
against the mesial margin of pm2. With the exclusion of
pm$5 and the apex of pm2, the left premaxillary teeth are
entirely hidden by the right ones. In the space between
right pm2 and pm3, the distal and mesial margins of two
adjacent left teeth can be seen: they are pm3 and pm4 or,
more probably, pm2 and pm3.

The right premaxillary teeth are closely set and are
spaced apart from the maxillary ones by a diastema, em-
phasised by the fact that the pm-m sutural surface does
not extend ventrally to the level of the tooth row, creat-
ing a shallow ventral concavity between the two bones,
similar to the so called “subnarial gap” (Rauhut, 2003;
Gauthier, 1986; Welles, 1984). A symmetric diastema is
visible in the left tooth row. A slightly enlarged 3" or 4°
dentary tooth, fitting the subnarial gap, is instead reported
in coelophysoids and Di/ophosaurus (Welles, 1984; Col-
bert, 1989; Rowe, 1989). A long premaxillary-maxillary
diastema associated with a very shallow gap appears to
be present, although not described, in the German Comp-
sognathus (Rauhut, 2003). According to Peyer (2006),
however, this space may be occupied by empty alveoli.
There is no evidence of a diastema between premaxilla
and maxilla in Sinosauropteryx, Sinocalliopteryx, the
French Compsognathus, Juravenator and Huaxiagna-
thus. Among other coelurosaurs, closely packed prema
xillary teeth followed by a diastema and a gap are pres-
ent in Dilong (Xu et al., 2004), whereas a diastema made
by a strong lateral excavation of the premaxilla interrupts
the otherwise continuous alveolar margin between the pre-
maxilla and the maxilla in Sinornithosaurus (Xu & Wu,
2001). In Scipionyx, the diastema left an empty space ros-
tral to ml, which permitted ml to be slightly procumbent.

In Scipionyx, pm2 and pmS are markedly larger than
pml, 3 and 4. Similarly, in Sinornithosaurus (Xu & Wu,
2001) pm2 is markedly larger than pm3 and pm4. Pm2
is the longest premaxillary tooth also in both Sihetun
specimens of Sinosauropteryx (NIGP 127586 and NIGP
127587), although it is not so markedly larger than pm3
and pm4.

The premaxillary teeth of Scipionyx are pointed and
feebly to moderately recurved especially toward the
apexes, where they are more recurved than the maxillary
teeth. On the contrary, in Juravenator the maxillary teeth
are more recurved than the premaxillary ones (GGhlich
& Chiappe, 2006). Pm1-4 are oval to suboval in cross-
section, and without serrations. PmS of Scipionyx is tran-
sitional in that it is not simply suboval in cross-section
(Fig. 45): on the lingual side, it presents two concavities
near the carinae, paralleling them. These concavities are
absent in the rostralmost teeth. A similarly shaped cross-
section is exposed at the broken end of the 2°‘ left max-
illary tooth, whereas from m3 on, the labiolingual com-
pression increases. In Compsognathus, the premaxillary
teeth are more rounded than the lateral teeth, and com-
pletely unserrated, too (Ostrom, 1978; Peyer, 2004). The

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 55

Stasi SA w*
o PA SI

Rae ©

Tenta,

Fig. 44 - A potentially diagnostic, and possibly autapomorphic, character of Scipionyx samniticus is the lower tooth row that is more
extended caudally than the upper tooth row. Heterodonty resembles that of other compsognathids. Note also the symmetrical devel-
opment of the tooth series in both maxillary rami, which is consistent with tooth-replacement having not yet started (see Ontoge-
netic Assessment). Left teeth are numbered in black, right teeth in white. Scale bar = 2 mm. See Appendix 1 or cover flaps for ab-

breviations.

Fig. 44 - La fila dei denti inferiori che si prolunga in direzione caudale più di quella dei denti superiori è un carattere potenzialmente
diagnostico e forse autapomorfico di Scipionyx samniticus. L’eterodontia ricorda quella di altri compsognatidi. Si noti anche lo svi-
luppo simmetrico della serie dentaria nei due rami mascellari, che indicherebbe che la prima sostituzione dei denti non era ancora
iniziata (vedi Ontogenetic Assessment). Denti sinistri numerati in nero, denti destri in bianco. Scala metrica = 2 mm. Vedi Appendice

1 o risvolti di copertina per le abbreviazioni.

premaxillary teeth of Compsognathus are also reported
(Peyer, 2004) to be clearly differentiated from the teeth of
the maxilla in having parallel mesial and distal borders on
the main body, and in being more columnar. The premax-
illary teeth are unserrated, expanded labiolingually, and
recurved towards the tips also in Sinosauropteryx (Currie
& Chen, 2001). There is no evidence of either a mesial or
distal carina or denticles in the rostralmost premaxillary
teeth of many other coelurosaurs, such as Caudipteryx (Ji
et al., 1998) and the basal deinonychosaurs Sinornitho-
saurus (Xu & Wu, 2001), Microraptor (Xu et al., 2000),
Byronosaurus (Norell e? a/., 2000) and Sinovenator (Xu
et al., 2002b). Distal serrations are instead present on the
penultimate premaxillary tooth of Juravenator (GOhlich
& Chiappe, 2006), whereas Sinocalliopteryx is reported
to have premaxillary teeth with small serrations on the

lingual side of the distal portions of the 1% and 2" ones,
and on the distal carina of the 4° one (Ji et a/., 2007a). In
basal theropods (Coelophysidae), both premaxillary teeth
and rostralmost dentary teeth are elliptical to nearly circu-
lar in cross-section, show little if any curvature and bear
few serrations to none at all (Tykoski & Rowe, 2004).

Maxillary teeth - The maxilla of Scipionyx bears 7
gently and regularly recurved, fang-like maxillary teeth.
A low number of maxillary teeth is reported also in the
other compsognathids, except Compsognathus, and in
a number of other small-sized theropods (for further
comparisons and comments about the possible signifi-
cance of this condition, see the Ontogenetic Assessment
section). The maxillary teeth of Scipionyx appear more
compressed labiolingually than the premaxillary teeth,

56 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 45 - Close-up of the pair of 5° premaxillary teeth of Scipionyx
samniticus. Scale bar = 0.5 mm. See Appendix 1 or cover flaps for
abbreviations.

Fig. 45 - Scipionyx samniticus, particolare del 5° paio di denti prema-
scellari. Scala metrica = 0,5 mm. Vedi Appendice 1 o risvolti di coper-
tina per le abbreviazioni.

with both labial and lingual sides slightly concave mesi-
odistally in proximity of the distal carina. As mentioned
above, ml is slightly procumbent, as a consequence of
the slightly upturned ventral border of the premaxillary
process of the maxilla. As in most theropods, the two
m4 are the largest teeth of the upper tooth row (Fig. 44).
In Scipionyx, they are twice the size of the largest den-
tary tooth and, thus, appear oversized with respect to the
whole tooth series. The last three teeth are the shortest
of the maxillary series and are preserved in their natural
position: because they are all inclined backwards, the
apex of each crown is at a level that is well-caudally be-
yond the caudal margin of its alveolus. As in the French
Compsognathus (Peyer, 2006), the maxillary tooth row
of Scipionyx extends back to the level of the rostralmost
tip of the jugal, but does not reach the caudal margin
of the antorbital fenestra. Contrary to the premaxillary
teeth, the maxillary teeth are well-spaced. The spacing is
greater than in Sinosauropteryx (Currie & Chen, 2001)
and the majority of theropods (e.g., Rauhut, 2003), in
which the teeth are relatively closely packed, but nar-
rower than in Sinocalliopteryx (Ji et al., 2007a: fig. 2),
in which every interdental space approaches two alveo-
lar diameters in length. AIl the maxillary teeth are pre-
served in place, fixed in their alveoli. The right crowns
are intact, with the exception of ml, the tip of which
is detached, rotated about 160° and exposed in lingual
view, and of m3 and m7, which lack their apexes. The

Fig. 46 - Close-up of the 3"° left maxillary tooth of Scipionyx samniti-
cus. The red arrow indicates the best preserved denticles. Scale bar =
0.5 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 46 - Scipionyx samniticus, particolare del 3° dente mascellare sini-
stro. La freccia rossa indica i denticoli meglio conservati. Scala metrica
= 0,5 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

left maxillary teeth are also intact, with the exception of
ml, which preserves few remains of the mesiolingual
surface, and m2, which lacks its apex.

Remarkably, the maxillary teeth are developed sym-
metrically in the right and left tooth rows (see Ontoge-
netic Assessment). The only exception is represented
by the broken right m3, of which the preserved FABL
(fore-aft basal length) is about 70% that of the left m3:
even if complete, the right m3 had to be smaller than the
left one.

In Scipionyx, serrations were previously observed on
the distal carina of m2 (Dal Sasso, pers. obs., 1997), but
are in fact clearly visible only in the teeth caudal to this
one. Serrations are present only on the distal carinae, even
if on the left m3 and m4 the mesial carinae appear faintly
crenulated rather than smooth. The serrations are com-
posed of simple denticles, perpendicular to the carinae,
with parallel sides and slightly rounded tips. On the distal
carina of the left m3, the 7 basal denticles are as long as
they are high; however, the apical denticles — beginning
with the 8° and 9', which are at mid-height of the crown
— become gradually shorter but maintain their length to
eventually form a crenulated margin in the apical third
of the crown. In m4, the denticles are clearly present up
to 3/4 of the height of the distal carina, but appear rather
shorter than those of m3. In mé and m7, the denticles are

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 57

inclined towards the apex, are higher than they are long
and are spaced more apart. The denticle count per mm,
measured on left m3, right m4 and right m6, ranges from
11 to 12. The denticle count per mm in the other comp-
sognathids is discussed in the Ontogenetic Assessment
section, as this character is potentially age-related. Con-
cerning the presence of the denticles, in Compsognathus
(Peyer, 2006) they are first recognised on the m4 and are
restricted to the upper third of the distal carina. The cau-
dalmost teeth, however, seem to be devoid of serrations.
In Huaxiagnathus, maxillary and caudal dentary teeth
have fine serrations on their distal carinae only (Hwang
et al., 2004). AIl maxillary teeth of Sinosauropteryx and
Juravenator bear denticles only on their distal carinae.
This pattern occurs also in more advanced coelurosaurs,
such as Microraptor (Xu et al., 2000) and Sinornithoides
(Russell & Dong, 1993), which, interestingly, represent a
basal dromaeosaurid and a basal troodontid, respectively.
Sinocalliopteryx differs from the other compsognathids,
including Scipionyx, in having caudal maxillary teeth that
are serrated on both carinae (Ji et a/., 2007a).

Dentary teeth - Ten well-exposed dentary teeth are
in place in the right dentary of Scipionyx. Most of the
teeth of the left dentary are hidden by the right hemiman-
dible, so no details on the lingual side of the crowns are
available: the mesial halves of left dl and d3 emerge just
rostrally to the crowns of the right dl and d3, whereas
only the tips of the left d4 and d5 emerge. Similarly to
the upper dentition, the lower tooth row presents a certain

degree of heterodonty: in fact, the first two rostral dentary
crowns are oval in cross-section, closely set and without
denticles; a third transitional crown bears denticles on the
distal carina and is flattened in the distal half; and the re-
maining caudal ones are moderately compressed laterally,
regularly well-spaced, serrated only on the distal carinae
and have a distal half of the labial side that is flat to slight-
ly concave (Figs. 47-48). Moreover, as in the maxillary
tooth row, the last two crowns of the mandibular series are
oblique and shorter, with tips directed caudally.

As in the maxillary teeth, the denticles are simple and
perpendicular to the distal carina, with parallel sides and
slightly rounded tips. In d3, the denticles are rather short
and present only in the basal half of the distal carina (Fig.
47). Caudally to this tooth, the denticles of the distal carina
become gradually higher, more widely spaced apart and
distributed along a longer portion of the carina, reaching
the apex in d7. From the basal half of d7, the height of the
denticles surpasses their length (Fig. 48) and they become
slightly inclined towards the apex of the crown.

The number of dentary teeth of Compsognathus is at
least 18 in the German specimen (Ostrom, 1978) and 21
in the French specimen (Peyer, 2006). Denticles are first
recognised at the level of the 10° tooth in the latter (Peyer,
2006). Sinosauropteryx has 12-15 dentary teeth and, as
in Scipionyx, the denticles appear from d3 on (Currie &
Chen, 2001). A relatively low number of dentary teeth is
reported also in a number of other small coelurosaurs (e.g.,
Sues 1977; Ji et al., 1998; Xu & Wu, 2001; Elzanowski,
2002).

Fig. 47 - Close-up of the left and right rostral dentary teeth of Scipionyx samniticus. The red arrow indicates the best preserved denti-
cles. Scale bar = 0.5 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 47 - Scipionyx samniticus, particolare dei denti rostrali destri e sinistri delle ossa dentali. La freccia rossa indica i denticoli meglio
conservati. Scala metrica = 0,5 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

58 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 48 - Close-up of the right caudal dentary teeth of Scipionyx samniticus. Note the concavity of the distal half of the labial side on each
tooth. The red arrows indicate the best preserved denticles. Scale bar = 0.5 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 48 - Scipionyx samniticus, particolare dei denti caudali dell’osso dentale destro. Si noti, su ogni dente, la concavità della metà
distale del lato labiale. La freccia rossa indica i denticoli meglio conservati. Scala metrica = 0,5 mm. Vedi Appendice 1 o risvolti di
copertina per le abbreviazioni.

In Scipionyx, 13 denticles per mm are counted on the
right d5 and the right d8, on which the serrations are best
preserved. The denticles are higher but narrower at the
base than the maxillary ones. A _well-preserved caudal
dentary tooth of the NIGP 127586 Sinosauropteryx speci-
men is 1.8 mm long and has 10 denticles per mm (Currie
& Chen, 2001). Seven denticles per mm are reported in
Huaxiagnathus (Hwang et al., 2004). In Sinocalliopteryx,
the teeth are reported to be serrated in the same way as the
maxillary ones, these having also mesial denticles. The
observations in the Ontogenetic Assessment section about
number of teeth and number of denticles per mm in the
maxillary tooth row are valid also for the dentary teeth.

Concerning the spacing of the crowns, Huaxiagnathus
resembles Scipionyx in having closely set rostral teeth and
well-spaced mid-caudal teeth (Hwang et a/., 2004: fig. 2).
Instead, the dentary teeth in Compsognathus (Peyer, 2006:
fig. 5) and Sinosauropteryx (Currie & Chen, 2001: fig. 1b)
are closely set along the entire tooth row; in Sinocalliopteryx
(Ji et al., 2007a: fig. 2) alternating large and small crowns,
regularly spaced along the entire tooth row, are present.

With respect to the maxillary crowns, the dentary
crowns are as short in Sinocalliopteryx as they are in Sci-
pionyx, but, even if they are similarly inclined caudally in
the two species, they are more recurved in Sinocalliop-
teryx. In Sinosauropteryx, the dentary teeth curve along
their whole length; in Compsognathus, the rear edge of
the tooth in lateral aspect is almost straight and perpen-
dicular to the jaw, whereas a marked kink is visible in the
apical third, where the edge is much recurved.

Finally, the dentition ofthe deinonychosaurs Microrap-
tor and Sinovenator (Hwang et al., 2002) is comparable
to that of Scipionyx only in having unserrated rostral teeth
and lateral teeth that are serrated only on the distal cari-
nae, the remaining characters differing markedly.

Heterodonty - Summing up, three morphotypes can be
identified in Scipionyx (Figs. 44-48). The first morphotype
consists in the rostralmost teeth (pm1-3 and dl), which have
crowns that appear rounded in cross section and are devoid
of carinae. A second morphotype is represented by the tran-
sitional teeth (pm$S, m1-2 and d2), which have crowns that
appear rounded in cross section, although to a lesser degree
than the former, and a concave surface close to the unser-
rated carina (with the mesial one displaced lingually). The

last morphotype includes the lateral teeth (from m3 and
d3 on), which have crowns that are more compressed la-
biolingually and that are more rounded in the mesial half
while being more flattened to concave in the distal half and
towards the serrated distal carina, which is in turn aligned
with the rostrocaudal axis of the tooth row.

The heterodonty of Scipionyx is comparable to that of
Sinosauropteryx (Currie & Chen, 2001), Huaxiagnathus
(Hwang et al., 2004), Compsognathus (Peyer, 2006; Os-
trom, 1978) and, possibly, Ornitholestes (Bakker, pers.
comm., 1998), in having rostral teeth that are rounded in
cross section and lacking denticles, and lateral teeth that
are more compressed labiolingually and serrated only on
the distal carina. Additionally, there is another feature of
Scipionyx that can be recognised in Sinocalliopteryx (Ji et
al., 2007a), Juravenator (Gòhlich & Chiappe, 2006) and
Compsognathus (Peyer, 2006): caudalmost teeth that have
not only the tip, but also the rest of the crown, curving
back continuously, and that are smaller and more triangu-
lar in lateral view. The variation that led Zinke (1998) to
distinguish three morphotypes referred to cf. Compsogna-
thus is also consistent with such a heterodonty.

In Scipionyx, the size of the rostral teeth is remarkable.
Some of the premaxillary teeth are comparable in size to
the largest maxillary teeth, whereas the rostral dentary teeth
are larger than the more caudal ones. This condition is dif-
ferent to that of the French Compsognathus, Huaxiagna-
thus and Sinocalliopteryx, in which the premaxillary teeth
are smaller than the maxillary teeth and the rostral dentary
teeth are as large as the lateral ones; in addition, it is defì-
nitely opposite to that of the tyrannosauroids, in which the
rostral teeth are markedly smaller than the lateral teeth.

Hyoid apparatus

The preserved elements of the hyoid apparatus con-
sist of a pair of slightly flattened, apically broadened and
gently curved rods, identifiable as ceratobranchials I (Fig.
49). They lie ventral to cervical vertebrae 2 and 6, and are
still parallel to each other but, contra Holtz et al. (2004),
fossilised caudally of the position they had in life, which
would have been in-between the caudal halves of the two
hemimandibles (Figs. 21-22). Although at first glance the
ceratobranchials I of Scipionyx seem disproportionately

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 59

large with respect to the posteranial skeleton, their size is
appropriate with the size of the skull: in fact, they mea-
sure only about half the length of the mandible in most of
the dinosaurs in which they are preserved (e.g., Compso-
gnathus [Peyer, 2006] and Proceratosaurus [Woodward,
1910; Rauhut et al., 2009]).

Contrary to a still persistent common belief, a number of
recent studies have demonstrated that ceratobranchials I are
not rarely found in association with dinosaur skulls (e.g.,
Weishampel et al., 2004). Among theropods, a list of the
taxa we encountered during this study comprises about 15
genera in which ceratobranchials I are anatomically în situ:
Jinfengopteryx (Ji et al., 2005), Sinosauropteryx (Currie
& Chen, 2001; Ji ef a/., 2007b), Compsognathus (Ostrom
1978; Peyer, 2006), Sinornithosaurus (Xu & Wu, 2001),
coelophysoids (Colbert, 1989; Rowe, 1989), Carnotaurus
(Bonaparte et al., 1990), Pelecanimimus (Pérez-Moreno et
al., 1994), Sinornithomimus (Kobayashi & Lii, 2003), Cau-
dipteryx (Ji et al., 1998), Shuvuuia (Chiappe et al., 2002),

Sinornithosaurus (Xu & Wu, 2001) and, possibly, Bambi-
raptor (Burnham et al., 2000). In the MGI 100/1001 speci-
men of Shuvuuia, they seem to run just below the pterygoi-
ds, almost appressed to their ventral margin and similarly
angled (Chiappe et a/., 2002: fig. 4.6B).

A displaced ceratobranchial I was described for Hua-
xiagnathus (Hwang et al., 2004), some tyrannosaurids
(RTMP 91.36.500) and some deinonychosaurs (Colbert
& Russell, 1969; Russell & Dong, 1993). Several authors
(e.g. Currie & Zhao, 1993a; Rauhut, 2003; Holtz et al.,
2004) reported ceratobranchial I in Sinraptor dongi. It
seems to us that the alleged ceratobranchial I of Sinraptor
(Currie & Zhao, 1993a: fig. 12A) does not match the shape
of any ceratobranchials I so far described (Weishampel e?
al., 2004); rather, it strongly resembles the supradentary-
coronoid of Daspletosaurus in having a pointed rostral
end and a distinctly twisted caudal end (Currie, 2003: fig.
35). Therefore, we regard the alleged ceratobranchial I of
Sinraptor as being a supradentary-coronoid.

POSTCRANIAL AXIAL SKELETON

With the exception of the missing portion of the tail,
all the elements of the axial skeleton of Scipionyx are pre-
served, although with slight to moderate post mortem dam-
age and deformation (Figs. 21, 22). The vertebral neural
arches are still in articulation and the zygapophyses inter-
connected. Some vertebral centra, especially those that lie
at the level of the cervicodorsal transition and in the sacral
region, are instead detached from their neural arch and
disarticulated from each other. This condition is probably
linked to the immaturity of the specimen (see Ontogenetic
Assessment) and to the cylindrical shape of the centra,
which favoured their mobility (see Skeletal Taphonomy).
The ribs, which are still in contact with their respective
vertebrae, and the gastralia, which lie almost undisturbed
and only slightly fractured by the diagenetic crushing, al-
low reconstruction of the chest and abdomen, and an esti-
mation of their respective volume (see Palaeobiology).

Vertebrae

In its preserved part, the vertebral column of Scipionyx
is composed of 37 elements: 23 presacrals, 5 sacrals and 9
caudals. The presacral vertebrae do not vary markedly in
length. As in most theropod dinosaurs, the morphological
transition of vertebrae from cervical to dorsal is gradual,
so much so that their regionalisation can be established
only tentatively, mainly basing assumptions on the most
apparent regionalisation of the ribs. Taking for certain that
there are 23 presacral elements, we consider the first 10
vertebrae to be cervicals, and the remaining 13 to be dor-
sals, on the basis of the features described below.

A similar vertebral count is reported in many basal
neotheropods (e.g., Tykoski & Rowe, 2004; O°Connor,
2007), in basal Tetanurae (Holtz et a/., 2004) and in
most compsognathids. Regarding the compsognathids,
Compsognathus (Peyer, 2006) has 23 presacral elements
(10 cervicals, 13 dorsals), most likely 5 sacrals and 30
preserved caudals; Sinosauropteryx (Currie & Chen,
2001) has 23 presacral elements (10 cervicals, 13 dor-
sals), probably 5 sacrals and 64 caudals; Huaxiagnathus

(Hwang et al., 2004) has 23 presacral elements (10 cer-
vicals, 13 dorsals), an indeterminate number of sacrals
(they are covered by the ilia) and 25 preseved caudals;
and Juravenator (Gòhlich & Chiappe, 2006) is reported
to have 21-23 presacrals (8-10 cervicals, 13 dorsals), an
uncertain number of sacrals and 44 preserved caudals.
The only exception may be represented by Sinocalliop-
teryx: Ji et al. (2007a) reported 11 cervicals, 12 dorsals,
an indeterminate number of sacrals and 49 caudals, but
in all likelihood this variation in the cervical count de-
pends on a different interpretation of a transitional ver-
tebra, considered as the last cervical rather than the first
dorsal. Even other non-maniraptoran coelurosaurs such as
the Tyrannosauroidea (Holtz, 2004) and the Ornithomi-
mosauria (Makovicky et al., 2004) have 23 presacrals (10
cervicals, 13 dorsals), although the former taxon presents
with a partially sacralised last dorsal, and the latter taxon
presents with 5 sacrals in the basal forms Shenzhousaurus
(Ji et al., 2003) and, probably, Archaeornithomimus (Ma-
kovicky et al., 2004), but tends to have 6 sacrals in more
derived forms. In turn, different and variable vertebral
counts characterise the derived coelurosaurs belonging
to the Maniraptora (see, for example, Weishampel ef al.,
2004; and references therein).

Cervical vertebrae - The very well-preserved cervi-
cal series emerges in right lateral view with a partial,
faint exposition of its dorsal aspect between C2 and C6.
The cervical vertebrae appear sharply angled in lateral
view, with the cranial articular surfaces elevated dorsal-
ly with respect to the caudal surfaces (Figs. 49-50). This
offset would have promoted the S-shaped curve typical
of most theropods (e.g., Carpenter et a/., 2005b; Naish
et al., 2001), observed when the articulated cervical col-
umn is viewed in lateral view (see also Skeletal Taphon-
omy). The cervicals of Scipionyx have low, rectangular
neural spines, prominent zygapophyses, overhanging
epipophyses and well-developed transverse processes
bearing diapophyses. All parapophyses, set on the cran-
ioventral corner of the centra, are hidden by the heads of
the cervical vertebrae.

60 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 49 - Cervical vertebrae, cervical ribs and ceratobranchials I of Scipionyx samniticus. Scale bar = S mm.
Fig. 49 - Vertebre cervicali, costole cervicali e ceratobranchiali I di Scipionyx samniticus. Scala metrica = 5 mm.

As in most theropods, the atlas of Scipionyx is com-
posed of the neurapophyses and the intercentrum. The
atlantal neurapophysis bears a well-developed, rounded
epipophysis, and a short, well-delineated postzygapophy-
sis. Compared to other theropods (e.g., Majungasaurus
[O”Connor, 2007]; Deinonychus [Ostrom, 1969]), the epi-
pophysis of Scipionyx seems rather short, as a consequence
of the marked dorsoventral development of the neurapo-
physis, but in any case it projects caudally to the postzy-
gapophysis. The atlas, aligned with the skull, lies in lateral
view, whereas the axis is in laterodorsal view. On account
of the slight rotation that occurred between the two ele-
ments, the atlantal postzygapophysis is not superimposed
anymore onto the prezygapophysis of the axis, leaving the
articular surface of the latter visible. In fact, contrary to all
other zygapophyses, which have articular surfaces lying
on the transverse plane, the contact between the atlantal
postzygapophysis and the axial prezygapophysis occurs
mainly in a sagittal plane. The atlantal intercentrum, which
is located ventrally to the axial prezygapophysis and cau-
dally to the caudal margin of the squamosal, probably lost
its contact with the neurapophysis, but the presence of the
prezygapophysis prevents us from verifying this. The ex-
posed portion is more triangular than rhomboidal in shape
and appears to be higher than wide.

Unlike the spines of the following cervicals, the axial
neural spine is rather tall and rounded, resembling those
of the basal ornithomimosaurs (Kobayashi & Lii, 2003;
Kobayashi & Barsbold, 2005). Similar to the other cer-
vical neural spines, it seems compressed mediolaterally
rather than flared transversely. The axial epipophysis is
large, not pointed, caudally directed and extended beyond
the postzygapophysis. Holtz ef al. (2004) indicate the
presence of axial prezygapophyses and well-developed
postzygapophyses bearing epipophyses in larger teta-
nurans, such as the tyrannosauroids, but they do not men-
tion the condition in small forms. According to Rauhut
(2003), the character “axis bearing well-developed epipo-
physes that extend beyond the postzygapophyses” is wide-
spread in Neotheropoda, but among coelurosaurs is lim-
ited to tyrannosaurids and deinonychosaurs (e.g., Norell

& Makovicky, 2004). In addition, axial epipophysis ex-
tending slightly more caudally than the caudal end of the
postzygapophysis is reported also in the ornithomimosaur
Sinornithomimus (Kobayashi & Li, 2003). In Scipionyx,
a possible axial pleurocoel (pneumatopore) is the depres-
sion visible just caudal to the base of the axial transverse
process (Fig. 51). In size and position, it resembles the
pleurocoel figured by Madsen (1976: pl. 11M) in A//osau-
rus. Such a pleurocoel is absent in coelophysoids, present
in Ceratosaurus and the abelisaurids (Tykoski & Rowe,
2004), and often present in the Tetanurae, with the excep-
tion, for example, of the megalosaurids (Rauhut, 2003;
Holtz et al., 2004). The centrum of the axis of Scipionyx is
about as long as the centra of the other cervical vertebrae.
Similar to its neural arch, also the axial centrum is slightly
rotated, appearing in laterodorsal view and exposing the
right neurocentral surface. The diapophysis has lost the
contact with the tuberculum of its rib, these elements ap-
pearing distant from each other. Just below the bifid head
of the axial rib, dorsal to the tuberculum and in a cranial
direction, emerges a portion of bone which we tentatively
refer to as the cranial margin of the axial intercentrum,
still connected to the centrum but not yet fused to it (Figs.
50B, 51). A possible sign of the odontoid process of the
axis, which is not directly exposed, is the relief under the
cranioventral margin of the neural arch and the atlantal
intercentrum.

The neural spine of C6 is exposed in dorsal view,
emerging vertically from the fossiliferous slab, and show-
ing the absence of a spine table (Fig. 52). The height of
the neural spines decreases from C2 to C4; from C5 to
C7 the neural spines are equally low but are also elongate
craniocaudally; from C8 to C10 they become higher but
shorter, as in C3.

In Compsognathus (Peyer, 2006), the neural spines are
long and very low throughout the entire cervical series. In
Sinosauropteryx (Currie & Chen, 2001), the cervical neural
spines are on average as low as in Scipionyx and Compso-
gnathus, but are craniocaudally shorter. Based on the pub-
lished figures, the neural spines of Juravenator (GG6hlich
& Chiappe, 2006) seem to resemble those of Scipionyx. In

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 61

atlantal neurapophysis
axial elements
cervical 3

cervical 4

cervical 6

cervical 7
cervical 8

cervical 9

cervical 5 Li cervical 10

atlantal intercentrum

LI axial intercentrum

indeterminate bone

calcite vein

Fig. 50 - Line drawings of the cervical vertebrae of Scipionyx samniticus. A) neural arches; B) centra, intercentra.
Fig. 50 - Disegni al tratto delle vertebre cervicali di Scipionyx samniticus. A) archi neurali; B) centri e intercentri.

Huaxiagnathus (Hwang et al., 2004) and Sinocalliopteryx
(Ji et al., 2007a), the spines are short, and comparatively
high and blade-like in the cranial cervicals, whereas they
are long and low in the caudal ones. Low neural spines
are reported also in other small-sized coelurosaurs such as
“Calamosaurus foxii” (Naish et al., 2001), Dilong (Xu et
al., 2004), Ornitholestes (in the alleged C4 and C6; the
spine of C3 is instead short and tall) and Coe/urus (in
which they extend as long as the arch [Carpenter et al.,
2005b]), as well as in the ornithomimosaurs (Kobayashi &
Barsbold, 2005; Makovicky et al., 2004), in Ngwebasau-
rus (de Klerk et a/., 2000) and in derived therizinosauroids
and oviraptorosaurs (K.irkland ef a/., 2005).

In all the post-axial cervical vertebrae of Scipionyx,
due to the extreme interlocking of the neural arches
caused by the backwards bending of the neck (see Skel-
etal Taphonomy), the postzygapophyses have slid on the
prezygapophyses up to the maximum limit of craniocau-
dal sliding permitted by their articular surfaces. Espe-
cially in the cranial and mid cervical arches, the articular
surfaces of the prezygapophyses, observed in lateral view,
are convex, with the cranial portion flexed ventrally. This
flexion gives an indication of the intervertebral mobil-
ity in the neck of Scipionyx (see Skeletal Reconstruction
And ...). The postzygapophyses are stout and rounded,
and slightly more pointed in the last three vertebrae. As

62 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 51 - Close-up of the atlas and the axis of Scipionyx samniticus.
Scale bar = 1 mm. See Appendix 1 or cover flaps for abbreviations.
Fig. 51 - Atlante ed epistrofeo di Scipionyx samniticus. Scala metrica
= 1 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

in Sinosauropteryx (Currie & Chen, 2001), in Scipionyx
the prezygapophyses are elongate at the front of the neck.
They tend to become gradually shorter at the back, es-
pecially between C8 and C10, resulting markedly shorter
than the postzygapophyses, which in turn remain constant
in length along the whole cervical series. Unlike Scipio-
nyx, in Huaxiagnathus the prezygapophyses are longer
than the postzygapophyses in the cranial cervicals, and
the latter apophyses become elongate at the back (Hwang
et al., 2004), as occurs in the basal ornithomimosaur
Sinornithomimus (Kobayashi & Lii, 2003). Also Comp-
sognathus (Peyer, 2006) has prezygapophyses that are
larger than those of Scipionyx, with a more rounded apex
than that of the postzygapophyses. In Scipionyx, both zy-
gapophyses extend far beyond the ends of their centra,
whereas in Sinosauropteryx this occurs only in the prezy-
gapophyses (Currie & Chen, 2001). Concerning the ven-
tral flexing of the prezygapophyses, according to Rauhut
(2003) the articular surface is more or less straight in non-
coelurosaurian theropods, whereas the cranial part of the
articular surface is flexed ventrally in Sinosauropteryx
(Currie & Chen, 2001), Ornitholestes (Carpenter et al.,
2005b), Nqwebasaurus (de Klerk et al., 2000), birds (e.g.,
Archaeopteryx), dromaeosaurids, ornithomimosaurs and
oviraptorosaurs (Makovicky, 1995). Unlike Scipionyx, in
Coelurus the prezigapophyses are not flexed in the cranial
cervical vertebrae, and become slightly flexed only in the
caudal ones.

In Scipionyx, right and left epipophyses can be seen
from CI to C8, projecting caudally to the postzygapo-
physes. Apart from the atlas, the vertebra which bears the
largest epipophyses is C6; in this vertebra, the apexes of
the epipophyses are pointed rather than rounded (Fig. 52).
From C6 to C10, the epipophyses become smaller, and in
C10 they are no longer clearly recognisable. In Compso-
gnathus (Peyer, 2006), small epipophyses are present and

Fig. 52 - Close-up of the neural arch of the 6" cervical vertebra
of Scipionyx samniticus. Its dorsal aspect shows the absence of a
spine table. Scale bar = 1 mm. See Appendix 1 or cover flaps for
abbreviations.

Fig. 52 - Scipionyx samniticus, particolare dell’arco neurale della 6°
vertebra cervicale. La norma dorsale mostra bene la mancanza di una
mensola spinale. Scala metrica = 1 mm. Vedi Appendice 1 o risvolti di
copertina per le abbreviazioni.

are placed distally on the postzygapophyses. Small epi-
pophyses are reported also in Huaxiagnathus (Hwang et
al., 2004). The epipophyses in Sinosauropteryx (Currie &
Chen, 2001) are represented only in the first few cervicals
of NIGP 127586 by low bumps and, unlike in Scipionyx,
they do not overhang the postzygapophyses caudally, rath-
er, they are oriented more dorsally. Epipophyses similar to
those of Scipionyx can be seen in the preserved cervicals
of Ornitholestes (Carpenter et al., 2005b), whereas small
epipophyses laterally connected and not overhanging the
postzygapophyses caudally are reported in Ngwebasaurus
(de Klerk et al., 2000).

Concerning the diapophyses, in Scipionyx the most ap-
parent change occurs between C2 and C4: in the axis the
diapophysis is small and supported by a transverse pro-
cess ventrally directed; in C3 it increases in size and the
transverse process is slightly raised albeit still ventrally
directed; in C4, the diapophysis is twice as large as in
C2 and the transverse process is longer and even more
raised, directed lateroventrally. This pattern is present in
most neotheropods, with the exception of some very bas-
al forms (e.g., Welles, 1984; Tykoski & Rowe, 2004) in
which the transverse processes and their diapophyses are
reduced or absent in C2. In Scipionyx, the transverse pro-
cesses continue to rise gradually from C6 to C10, expos-
ing more clearly the articular surfaces of the diapophyses.
The transverse processes become horizontal in the cra-
nialmost dorsal vertebrae (see below).

Along the cervical series, parallel to the shortening of
the arch that affected primarily the prezygapophyses, the
transverse processes change position also in a craniocau-
dal direction: they are positioned about midway between
the zygapophyses in the cranial and mid cervicals; cau-
dally, they gradually move closer to the prezygapophyses.
The lateral protrusion of the transverse processes high-
lights the presence of prezygodiapophyseal and postzy-
godiapophyseal laminae, which connect each transverse
process with the zygapophyses of that side; a prezygoepi-
pophyseal lamina (sensu O°Connor, 2007) can be seen,
too (Figs. 53-54).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 63

Fig. 53 - Shaded drawing of the 4'* cervical vertebra and ribs of Scipionyx samniticus. See Appendix 1 or cover flaps for

abbreviations.

Fig. 53 - Disegno ombreggiato della 4* vertebra cervicale di Scipionyx samniticus e delle sue costole. Vedi Appendice 1 o risvolti di

copertina per le abbreviazioni.

Because of the strict interlocking of the vertebrae, it
is difficult to see whether the cervical centra are amphi-
platyan or opistocoelous. However, thanks to a slight up-
lift of the centrum of C6, the cranial articular surface of
this vertebra can be observed. It appears moderately con-
vex, indicating that the cervical vertebrae of Scipionyx are
moderately opisthocoelous. Thus, contra Peyer (2006),
the cervical vertebrae of Scipionyx are not biconcave.
The cervical centra are opisthocoelous also in Compso-
gnathus (Ostrom, 1978; Peyer, 2006) and, apparently, in
Sinocalliopteryx (Ji et al., 2007: fig. 4a). Sinosauropteryx
is reported to have biconcave cervical vertebrae (Currie
& Chen, 2001).

In “Calamosaurus foxii” (Naish et al., 2001) and Orni-
tholestes the cervical centra are strongly opisthocoelous,
whereas they are amphicoelous in Coel/urus (Carpenter ef
al., 2005b). Among basal tyrannosauroids, the cervical
vertebrae are opisthocoelous in Di/ong (Xu et al., 2004)
and amphicoelous in Guanlong (Xu et al., 2006), the latter
being usually considered to be more basal than the former.
Strongly opistocoelous vertebrae are known also in some
non-coelurosaur tetanurans, such as the Spinosauroidea
and the Allosauroidea (Rauhut, 2003). Opisthocoelous
cervical vertebrae are also known in a few Maniraptora,
such as the alvarezsaurids (Perle e? a/., 1993; Chiappe et
al., 1998), even if most of the maniraptoran theropods, in-
cluding the basal alvarezsaurid Ngwebasaurus (de Klerk
et al., 2000), tend to have amphicoelous or amphiplatyan
centra (e.g., Ji ef al., 1998; Norell & Makovicky, 2004;
Xu & Norell, 2004; Kirkland et a/., 2005).

In Scipionyx, the centra of the caudal cervical verte-
brae are taller than those of the cranial ones. As in the vast
majority of theropods, including other compsognathids
(e.g., Currie & Chen, 2001; Ji et al., 2007a), in lateral
view the centra are concave ventrally, with articular faces

Fig. 54 - Close-up of the 5‘ and 6' cervical vertebrae (from right to left)
and the ends of the preceding ribs of Scipionyx samniticus. Scale bar
=2 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 54 - Scipionyx samniticus. Quinta e 6° vertebra cervicale (da destra
a sinistra) e terminazioni delle costole precedenti. Scala metrica=2 mm.
Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

64 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 55 - Dorsal vertebrae and dorsal ribs of Scipionyx samniticus. Scale bar = 5 mm.
Fig. 55 - Vertebre e costole dorsali di Scipionyx samniticus. Scala metrica = 5 mm.

parallel but inclined with respect to the long axis of the
centrum, not perpendicular, so that in lateral aspect the
cranial face is offset well-above the caudal face. The ar-
ticular surfaces become less oblique towards the end of
the cervical series, becoming nearly vertical in the 2°° dor-
sal centrum. In Scipionyx, the length of the cervical centra
is on average 5/2 their height; the length of the centra is
three times their height in Compsognathus (Peyer, 2006).
Thus, the cervical centra of the compsognathids are lon-
ger than those of juvenile tyrannosaurines and other ty-
rannosauroids, in which they are as long as or longer than
they are tall (Holtz, 2004; Xu et al., 2004), but shorter
than those of the therizinosauroids (e.g., Zanno, 2010).
In Scipionyx, the centra of C6 and C7 are slightly more
elongate craniocaudally than the others. This is similar to
Sinosauropteryx (Currie & Chen, 2001) , in which the 6"
to 8! cervicals are longer than the cranial ones. The cau-
dalmost cervical centra of Scipionyx are not significantly
longer than the cranial dorsal elements.

The neurocentral suture is visible on all the cervical
vertebrae (see Ontogenetic Assessment). The centra are
still connected to their respective arches only in C4 and
G6stin CGS TCSEC7C8 and C9scentratandiarenestane
so separated, especially in the caudalmost ones, that the
dorsal surface of the centra is often visible.

Small pneumatopores can be seen in the middle of the
centra of C3-C5, in the form of single pnreumatic foramina
(Figs. 53-54). As previously reported by Rauhut (2003;
and references therein) and Wedel (2009), in modern birds
the pneumatisation ofthe cervical vertebral column during
ontogeny starts at the cervical-dorsal transition and then
continues cranially and caudally, and in non-avian thero-
pods the pneumatisation of the vertebral column probably
followed the same ontogenetic pattern. Thus, in Scipionyx
the finding of the foramina in the cranial half of the cervi-
cal series but not in the caudal half is quite unexpected.
Single pleurocoels on each side of the centrum have been
described for Compsognathus (Peyer, 2006), in which
they are barely visible, opening just caudal and slightly

ventral to the diapophyses; in Sinosauropteryx (Currie &
Chen, 2001), in which, as in Scipionyx, they open in the
cranial half of the series, and more precisely caudodorsal
to the parapophysis of the 5'® cervical of NIGP 127586,
and in the 3'° and 4° cervicals of NIGP 127587; in Jurave-
nator (Gòhlich & Chiappe, 2006); and more generally in
basal Tetanurae (Holtz et al., 2004), including Dilong (Xu
et al., 2004) and Nqwebasaurus (de Klerk ef al., 2000).
No pleurocoels are present in Ornitholestes, whereas they
vary in size and number in Coelurus (Carpenter et al.,
2005b). Two foramina penetrating the lateral surface of
the centrum are reported for neoceratosaurs (Tykoski &
Rowe, 2004), possibly for the basal tetanuran Fukwirap-
tor (Azuma & Currie, 2000), and, according to Holtz et al.
(2004), for the juvenile tyrannosauroid Shanshanosaurus,
some adult therizinosauroids and oviraptorosaurs, and the
troodontid 7roodon.

In Scipionyx, the centrum of C10 is detached from the
neural arch and has moved ventrally for a distance almost
equal to the height of the centrum itself. It has rotated on
its craniocaudal axis and has become exposed in a full
dorsal view, showing both neurocentral surfaces for their
whole length. This view has permitted us to note that the
transverse diameter of the centrum is only slightly less
than (almost 3/4) its craniocaudal diameter. This vertebra
can be considered a transitional vertebra (cervicodorsal),
as its rib possesses tuberculum and capitulum that are
larger and stouter than those of the preceding cervicals.

Dorsal vertebrae - The boundary between cervical
and dorsal vertebrae is defined by an abrupt increase in
rib size, affecting especially the capitula. So, as pointed
out by Dal Sasso & Signore (1998a), we consider the first
dorsal of Scipionyx as being the vertebra immediately
successive to the one in which the centrum is completely
separated from the neural arch (see above). That vertebra,
in fact, has a stout diapophysis that is still articulated to
its rib through an even stouter tuberculum; the capitulum,
although fractured by a calcite vein crossing it, appears

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 65

Q
o)
D7)
d
N

dorsal 6

dorsal 8

dorsal 9

dorsal 10

dorsal 11

dorsal 12

dorsal 13

indeterminate bone

calcite vein

Fig. 56 - Line drawing of the dorsal vertebrae of Scipionyx samniticus. A) neural arches; B) centra.
Fig. 56 - Disegni al tratto delle vertebre dorsali di Scipionyx samniticus. A) archi neurali; B) centri.

larger than the preceding ones. At mid-length, however,
this rib tapers abruptly, terminating in a thin shaft compa-
rable in size to the cervical ribs. Therefore, the 15 dorsal
rib (Drl) morphologically represents a transitional rib.
The 13 dorsal vertebrae of Scipionyx are faintly to mod-
erately constricted, not spool-shaped, with neural spines
that are low and craniocaudally expanded to an increased
degree, and without epipophyses (Figs. 55, 56).

The dorsal vertebrae have maintained anatomical con-
nection and exposure in right lateral view only along the
neural arches. D2, although partially covered by the large
dichocephalous heads of the right Dr2 and Dr3, is well-

outlined in lateral view, as the centrum and the neural arch
are still in articulation. In D2, part of the cranial articular
face is also exposed, appearing moderately convex. In
DI and from D3 to DI], the centra have detached from
their arches in various ways. In DI, D4 and D$, which
lie in almost lateral view, the right neurocentral surface
can be seen. Regarding D3, the cranial articular face of its
centrum, labelled “h” (humerus) in Dal Sasso & Signore
(1998a: fig. 2), is well-exposed. This face, which has ro-
tated not only laterally up to expose the cranial surface,
but also anticlockwise as far as to point its ventral mar-
gin towards the skull, is convex just like D2. D5, D6, D7

66 CRISTIANO DAL SASSO & SIMONE MAGANUCO

and D9 are exposed in dorsal view, revealing both neuro-
central sutural surfaces, and the floor of the neural canal,
which appears rather broad, in particular at the level of the
pectoral girdle (see also S5, below). In addition, the centra
of D6 and D7 have rolled ventrally into the rib cage. The
centrum of D6 shows its articular caudal face, but unfor-
tunately it is deformed and consequently not informative.
The centrum of DI0 also lies in dorsal view, but it exposes
only the left neurocentral sutural surface, as the rest of the
centrum is covered by the ribs and the intestine. DI1 lies
in an odd position: the centrum has rotated laterally about
a right-angle up to expose the cranial articular face, which
is oval just like D3 (5/4 wider than high). At this level
of the dorsal series, the articular faces are almost flat, as
for the centra of D12 and D13, which in lateral view are
slightly spaced from each other. In other compsognathids,
the dorsal vertebrae have been described as platycoelous
(Peyer, 2006) or amphiplatyan (Hwang et al., 2004), with-
out distinction between cranialmost and caudalmost cen-
tra, whereas they are reported to be amphicoelous in some
ornithomimosaurs (e.g., Kobayashi & Barsbold, 2005).
D12 and D13 are the dorsal vertebrae most clearly ex-
posed in Scipionyx, and they still possess a neural arch
that is articulated but unfused to the centrum, with the
base of the arch forming a bilobed outline marked by a
median concavity.

The apexes of the neural spines of the dorsal verte-
brae appear faintly expanded transversely. Although the
neural arches are exposed mainly in lateral view, these
faint transverse expansions are visible at the top of D2,
D3 and D6 thanks to a slight distortion of the spine (Fig.
57). Such structures cannot be considered as true spine
tables: they simply represent the reinforced, rugose dorsal
margins of the spines.

The cranial dorsal vertebrae bear relatively small,
trapezoidal neural spines. They are the craniocaudally
shortest of the series, being comparable in length to the
last cervicals. In Scipionyx, beginning from DA, the neural
spines arise on the caudal half of the vertebrae, slightly
overhanging the succeeding vertebra. A similar pattern
can be seen in Compsognathus (Peyer, 2006), but already
from the 1% dorsal.

Fig. 57 - Close-up of the neural arches of the 2"° and 3" dorsal vertebra
(right to left) of Scipionyx samniticus, showing faint transverse expan-
sions going to the left (red arrows) and the right (green arrows) at the
top of the neural spines, but no true spine tables. Scale bar = 1 mm.
Fig. 57 - Scipionyx samniticus. Particolare degli archi neurali della 2°
e 3° vertebra dorsale (da destra a sinistra), che mostrano deboli espan-
sioni trasversali, a sinistra (frecce rosse) e a destra (frecce verdi) della
sommità delle spine neurali, ma non vere e proprie mensole spinali.
Scala metrica = 5 mm.

Unlike non-coelurosaurian tetanurans (e.g., A/losau-
rus) and some coelurosaurs, such as tyrannosauroids
(Carpenter ef al., 2005a) and ornithomimosaurs (Ko-
bayashi & Barsbold, 2005; Kobayashi & Lii, 2003), in
which the neural spines are relatively high, the spinous
processes increasing evidently in height throughout the
column towards the sacrum (Holtz et a/., 2004), in Scipio-
nyx the dorsal neural spines are relatively uniformly low
and shorter than the centrum, as is the case in Sinosau-
ropteryx (Currie & Chen, 2001). In the Italian compsog-
nathid, the caudal dorsal neural spines are also definitely
larger and craniocaudally more expanded than the cranial
dorsals, especially at mid-top. A similar craniocaudal ex-
pansion is found in the other compsognathids (Ostrom,
1978; Currie & Chen, 2001, Hwang et al., 2004; Gohlich
& Chiappe, 2006; Peyer, 2006) and was first described as
“fan-shaped” (Ostrom, 1978). Subsequently, the term was
applied to a variety of forms in literature. As reported by
Holtz ef al. (2004), fan-shaped dorsal neural spines are
described also in the basal dinosauromorph Marasuchus
and in the basal troodontid Sinovenator. Later, they were
found also in the troodontid Mei (Xu & Norell, 2004). In
Marasuchus (Sereno & Arcucci, 1994) and Sinovenator
(Xu et al., 2002b), the base of the spine is narrower than
the top in lateral view, and the top is slightly convex. The
caudal dorsal neural spines are longer craniocaudally at
the tip than at the base also in E/aphrosaurus (Rauhut,
2003), Mirischia (Naish et al., 2004), Huaxiagnathus
(Hwang et al., 2004), Sinocalliopteryx (Ji et al., 2007a)
and Sinosauropteryx (Chen et al., 1998; Currie & Chen,
2001). The neural spine is craniocaudally expanded also
in Scipionyx and Compsognathus (Peyer, 2006: fig. 2E),
but in these two taxa the point of maximum expansion is
at mid height, emphasised by the presence of beak-like
ligament attachments (see below). As a consequence, the
spine appears as a flattened hexagon, inclined caudally.
Given these differences, generically fan-like shaped
spines should not be considered as a character shared by
compsognathids, but rather the simultaneous presence of
relatively low spines that are significantly expanded cra-
niocaudally, in particular at the mid-top/top, with cranio-
caudally flat dorsal margins. This scenario is complicated
by the fact that, as reported by Naish et a/. (2004), the
degree of ossification of the dorsal tip and of the cranial
and caudal surfaces of the neural spine may increase dur-
ing ontogeny.

A peculiar feature of the dorsal vertebrae of Scipionyx
is the presence of beak-like extensions on the cranial and
caudal margin of the neural spine, just below the apical
margin. They are emphasised dorsally by concavities that
mark the craniodorsal and caudodorsal margins of the
spine. The cranialmost beak-like extension in Scipionyx
is preserved on the caudal margin of D6; more clearly
exposed are the ones on the cranial and caudal margins
of the best preserved neural spines, D9 and DII (Fig.
58). These extensions were interpreted as hyposphenes-
hypantra by Dal Sasso & Signore (1998a), but their prox-
imity to the top of the spine and the strict interlocking of
the pre- and postzygapophyses, which eventually oblit-
erate other accessory articulations, allow us to exclude
that interpretation. Very similar “hook-like” extensions
serving as attachments for the interspinal ligaments were
first described by Peyer (2006), who considered them as
diagnostic of compsognathids. Peyer (2006) noted their

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 67

Fig. 58 - Close-up of the neural arch of the 11° dorsal vertebra of
Scipionyx samniticus. Note the craniocaudally developed neural spine,
which is further expanded by the presence of beak-like attachments for
the ligaments. Scale bar = 1 mm. See Appendix 1 or cover flaps for
abbreviations.

Fig. 58 - Particolare dell’arco neurale dell’ 11° vertebra dorsale di Sci-
pionyx samniticus. Si noti la spina neurale sviluppata craniocaudal-
mente, ulteriormente espansa da attacchi di legamenti a forma di becco.
Scala metrica = 1 mm. Vedi Appendice 1 o risvolti di copertina per le
abbreviazioni.

presence in Ornitholestes, Compsognathus, Huaxiagna-
thus, Scipionyx and Sinosauropteryx. Similar structures
are also present in Di/ophosaurus, where they are called
“anterior and posterior shoulders” (Welles, 1984). Simi-
lar to Scipionyx, in Compsognathus “hook-like” ligament
attachments on the cranial border of the spine can be most
easily recognised on mid-caudal dorsal neural spines, in
particular in D7, D8, D9 and DII (Peyer, 2006).

The zygapophyses of the cranial and mid-dorsal ver-
tebrae extend less beyond the ends of their centra than in
the cervical vertebrae. In the caudal dorsals (e.g., D12)
the postzygapophyses do not extend caudally to the cen-
tra at all. The articular surfaces are hardly visible in all
zygapophyses except those between D7 and D8. Here, the
postzygapophysis of D7 and the successive prezygapo-
physis of D8 are not strictly in contact and show horizon-
tal articular surfaces.

The transverse processes, visible in most of the dor-
sals, appear rather short and, especially in the caudal half
of the series, slightly oriented caudally. According to
Holtz et al. (2004) and Peyer (2006), the shortness of the
transverse processes is common among Compsognathi-
dae. In Scipionyx, the orientation of the transverse pro-
cesses bearing diapophyses changes abruptly from ven-
trolateral to lateral between DI and D2. The orientation
is strongly dorsolateral in D7, dorsolateral in D9-DI11 and
again lateral in D12-D13. Most of the dorsal transverse
processes lie, therefore, more or less on the transverse
plane, as in Compsognathus, Huaxiagnathus and more
derived maniraptorans (Peyer, 2006).

In DI (Figs. 59-60), one can see that the diapophysis
is clearly buttressed by various pronounced laminae: in a
cranial direction by a prezygodiapophyseal lamina, and in
a caudal direction by a postzygodiapophyseal and a pos-
terior centrodiapophyseal lamina (sensu Wilson, 1999). In

Fig. 59 - Close-up of the neural arch of the 1% dorsal vertebra of Scipio-
nyx samniticus. Scale bar = 1 mm.

Fig. 59 - Particolare dell’arco neurale della 1° vertebra dorsale di Scipio-
nyx samniticus. Scala metrica = 1 mm.

some caudal dorsals (D9, D10, D12 and D13), the cra-
nial laminae, such as the prezygodiapophyseal and the
paradiapophyseal lamina, as well as the fossa delimited
by them, are more clearly recognisable; in D12, the prezy-
goparapophyseal lamina can be seen, too (Fig. 61). In DI,
a foramen on the bottom of the intrapostzygapophyseal
fossa is certainly present (Fig. 62), just like in Sinraptor
(Currie & Zhao, 1993a).

As is typical of nonavian theropods, the dorsal verte-
brae of Scipionyx exhibit a dorsal migration of the parapo-
physis from the cranioventral margin of the centrum to
the centre of the neural arch, moving caudally through
the series. The wide divergence between tuberculum and
capitulum of Dr4 indicates that, although not directly ob-
servable, in D4 the migration of the parapophysis on the
neural arch has not yet occurred. Similar observations
on the successive ribs (see below) suggest that the mi-
gration is completed at the level of D6-D7. The migrated
parapophyses are directly visible on D9, D10, D12 and
D13. More precisely, on D9 and DI0 they are located at
the cranial base of the arch; on D12, the parapophysis has
almost reached the height of the diapophysis; on D13, it
is perfectly aligned to the latter (Fig. 56A). At the same
time, besides migrating dorsocaudally through the series
and approaching the diapophyses, the parapophyses of
the caudal dorsal vertebrae are gradually reduced in size
(see also Ribs). Unlike Scipionyx, in most tetanurans the
parapophysis remains below the level of the transverse
process even in the caudalmost rib-bearing dorsal verte-
bra (Rauhut 2003).

In Scipionyx, the caudal dorsal centra are slightly
longer than the cranialmost ones. The centrum length
increases with each additional segment (see Table 1), as
occurs also in Compsognathus (Peyer, 2006). The latter,

68 CRISTIANO DAL SASSO & SIMONE MAGANUCO

ve

sis
sr ss

TÀ A
AL

‘N

16

#È,

di
O

Fig. 60 - Shaded drawing of the 1% dorsal vertebra and the 1% right dorsal rib of Scipionyx samniticus. See Appendix 1 or cover flaps

for abbreviations.

Fig. 60 - Disegno ombreggiato della 1° vertebra dorsale e della 1° costola dorsale destra di Scipionyx samniticus. Vedi Appendice 1 o

risvolti di copertina per le abbreviazioni.

however, differs from Scipionyx in having the 2°° dorsal
vertebra with a centrum much shorter than the cervical
ones, whereas in Scipionyx the cranial dorsals and the
cervicals are comparable in length. The dorsal centra of
Scipionyx are almost parallelepiped-like in shape. Unlike
allosaurids, which have a prominent constriction at mid-
section of the centra (Holtz et a/., 2004), and unlike Sino-
sauropteryx, which has deeply concave sides (Currie &
Chen, 2001), the dorsal centra of Scipionyx have almost
parallel sides when seen in dorsal view and in lateral view.
As in Sinosauropteryx (Currie & Chen, 2001), in lateral
view the concavity of the ventral edge is almost absent in
Scipionyx, resulting in it being even less marked than in
Compsognathus (Peyer, 2006). The ventral edges of the
cranial dorsal centra are mostly hidden by ribs (D1) and
muscular tissue remains (D2). Where the edges emerge,
ventral keels and hypapophyses are not visible. In taxa
that have well-developed keels and hypapophyses, they
are already reduced at the level of D3 (Rauhut, 2003).
Thus, in Scipionyx the absence of these structures in the
cranial face of the centrum of D3 is expected, and not in-
formative of the condition of the preceding vertebrae. Hy-

papophyses are absent in Compsognathus (Peyer, 2006).
They are present in the cranialmost dorsals of sinraptorids
and most Maniraptoriformes (Rauhut, 2003), but not in
the basal ornithomimosaur Garudimimus (Kobayashi &
Barsbold, 2005).

In the mid-caudal dorsal vertebrae of Scipionyx, the
craniocaudal diameter is 5/3 the dorsoventral diameter,
and 5/4 the mediolateral diameter. The caudal dorsal cen-
tra are similarly elongate in Mirischia (Rauhut, 2003: fig.
27C), on average twice as long as they are tall in Comp-
sognathus (Peyer, 2006) and Huaxiagnathus (Hwang et
al., 2004), and significantly even more elongated in dilo-
phosaurids, coelophysoids and ornithomimosaurs (e.g.,
Kobayashi & Lii, 2003), whereas they are as high as
long in Sinraptor, Allosaurus, Majungasaurus (Rauhut,
2003) and Sinosauropteryx (Currie & Chen, 2001), or
even higher in 7yrannosaurus and Deinonychus (Rauhut,
2003). Contrary to Scipionyx, in many coelurosaurs such
as Coelurus, Ornitholestes (Carpenter et al, 2005b),
Tanycolagreus (Carpenter et al., 2005a) and Garudimimus
(Kobayashi & Barsbold, 2005), the centra are reported to
be taller than they are wide.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 69

Fig. 61 - Close-up of the 12* dorsal vertebra of Scipionyx samniticus.
Scale bar = 1 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 61 - Dodicesima vertebra dorsale di Scipionyx samniticus. Scala me-
trica = 1 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

On the exposed surfaces of the dorsal vertebral centra
of Scipionyx, no pleurocoels are visible. This is expected,
as, according to Holtz ef a/. (2004), the compsognathids
differ from most of the other tetanurans in lacking any
dorsal pleurocoel. The condition in Scipionyx, however,
might be influenced by two factors: ontogeny and pres-
ervation. Concerning ontogeny, as mentioned above, ac-
cording to Wedel (2009) the pattern of pneumatisation of
the vertebrae in the Saurischia seems to reflect that of the
extant avian forms, with the phylogenetic trajectory with-
in the group matching what occurs during bird ontogeny.
The pneumatisation of the cranial dorsals could eventu-
ally have occurred later in Scipionyx, given that, although
not a basal theropod, the specimen is a very immature in-
dividual (see Ontogenetic Assessment). Concerning pres-
ervation, the exposure in Scipionyx of the lateral surface
of the centra dorsal and caudal to the parapophyses, where
the pleurocoels are usually found, is, unfortunately, very
limited, rendering confirmation of their absence difficult.
However in D2, which is well-exposed, no pleurocoel can
be seen. It is interesting to note that dorsal pleurocoels are
entirely absent also in some other taxa that have cervical
pleurocoels, such as E/aphrosaurus, Avimimus (Rauhut,
2003), Coelurus (Carpenter et al., 2005b) and the above
mentioned Comspognathus (Peyer, 2006) and Sinosaurop-
teryx (Currie & Chen, 2001), although in the latter genus a
problem of preservation cannot be ruled out. In addition,
in the Ornithomimosauria (Makovicky et al., 2004) and

Fig. 62 - Shaded drawing of the 9% dorsal vertebra and ribs of Scipionyx samniticus. See Appendix 1 or cover flaps for abbreviations.
Fig. 62 - Disegno ombreggiato della 9° vertebra dorsale di Scipionyx samniticus e sue costole. Vedi Appendice 1 o risvolti di copertina

per le abbreviazioni.

70 CRISTIANO DAL SASSO & SIMONE MAGANUCO

in the Troodontidae (Currie & Dong, 2001, Makovicky
& Norell, 2004; Xu & Norell, 2004) only the first one
or two dorsal centra are sometimes pneumatic to varying
degrees, whereas basal tyrannosauroids (Xu et al., 2004;
Carpenter ef al., 2005a; Xu et al., 2006) and Microrap-
tor (Hwang et al., 2002) are reported to lack pleurocoels/
pneumatic foramina.

In Scipionyx, the last dorsal vertebra (D13) lacks ar-
ticulated ribs. However, as both diapophysis and parapo-
physis are present on its right side, a Dr13 was in all like-
lihood present. Usually, the last dorsal rib is small and
short in theropods. The bony fragment preserved ventral
to the centrum may represent a splinter of the right Dr13,
lost during preparation, or of the left one, which would
emerge between the centrum of the D13 and the intestine
(see Ribs).

Sacral vertebrae - In Scipionyx, elements belonging
to 4 different sacral vertebrae are exposed (Figs. 63-64).
S1 is in continuity with the dorsal series, occupying the
position it occupied in the undisturbed skeleton. A small
portion of a successive centrum, representing S2, can be
seen in a space between the elements of the pelvic girdle.
The neural arches of the other two exposed vertebrae pre-
cede the caudal series. The space between the neural arch
of S1 and the penultimate arch matches in size the length
of two neural arches. Similarly, based on the length of
the centrum of SS, 5 elements are needed to fill the space
between the last dorsal and the 1%" caudal vertebrae. Thus,
the sacrum of Scipionyx is composed of 5 vertebrae, and
the arches preceding the caudal ones belong to S4 and S5.

Five sacrals (1 dorsosacral, 2 sacrals and 2 caudosacrals)
are the usual count for Neotheropoda (Tykoski & Rowe,
2004), and, as mentioned above, this is the number of
sacrals preserved in Compsognathus (Peyer, 2006). The
sacral count is unknown in Huaxiagnathus (Hwang et al.,
2004), Sinocalliopteryx (Ji et al., 2007a) and Juravenator
(GGhlich & Chiappe, 2006).

SI is the only sacral vertebra in which the neural arch
is still adjacent to its centrum, with the neurocentral su-
ture only slightly open. Because of a slight anticlockwise
rotation of the iliac blade, most of the right lateral as-
pect of SI is exposed ventral to the ventral margin of
the preacetabular ala of the ilium, and cranial to the con-
tact between the ilium and the pubis. The centrum does
not show pneumatopores, as for example in Guanlong
(Xu et al., 2006), and the neural arch, well-dorsal to the
neurocentral suture, bears an evident bumpy transverse
process. As the latter terminates laterally with a single
rugose facet, and no scars are visible on the centrum, we
infer that the transverse process and the rib attachments
were conjoined in SI. This condition resembles the one
of Tyrannosaurus, where a single, rugose, bifid facet is
seen on a process of the neural arch of S1 (Brochu, 2003)
not invading the centrum with its ventral part, contrary
to what occurs in A//osaurus (Madsen, 1976). Unfortu-
nately, in Scipionyx the transverse process facets and rib
facets of the successive sacral vertebrae are not exposed,
so it is impossible to establish if they are separated and
located dorsolaterally and lateroventrally across the neu-
ral arch and centrum, as occurs in the above mentioned,
well-known, large genera.

Fig. 63 - Sacral vertebrae and ribs of Scipionyx samniticus. Scale bar = 5 mm.
Fig. 63 - Vertebre e costole sacrali di Scipionyx samniticus. Scala metrica = 5 mm.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY TA |

In Scipionyx, the cranial margin of the centrum of S1
is perfectly aligned to the cranial margin of the iliac blade,
Just like in the French Compsognathus (Peyer, 2006). The
exposed tract of the ventral margin appears almost flat.
Regarding S2, two small portions emerge: a triangular
area, probably corresponding to the craniodorsal portion
of the centrum, is visible between the lesser (=anterior)
trochanter of the right femur and the pubic peduncle of the
right ilium; and the caudodorsal portion of the centrum
and the caudoventral base of the neural arch, are visible
between the lesser trochanter of the right femur and the
cranioventral margin of the acetabulum.

Nothing can be said about $3, as neither the neural arch
nor the centrum are exposed. It can be hypothesised that
this vertebra is still in place, sandwiched in-between the
ilia. In contrast, the elements forming S4 and S5 emerge
dorsal and ventral to the postacetabular ala of the right
ilium, due to the fact that they are disarticulated and have
lost their connection with the ilia. The degree of compres-
sion and deformation that occurred in this area of the skel-
eton renders it difficult to interpret elements and struc-
tures belonging to S4 and S5, and to establish in which
aspect they became fossilised (Fig. 64). The neural arch
of S4, from the top of the spine to its base, can be seen in
right lateral view. The ventral margin of the spine leans

MH sacra

MD sacra? °° sacral5 ( sacralrib

MI sacra 4 EI indet bones

Fig. 64 - Line drawings of the bones illustrated in Fig. 63. A) neural
arches and ribs; B) centra and indeterminate elements. See Appendix 1
or cover flaps for abbreviations.

Fig. 64 - Disegni al tratto delle ossa illustrate in Fig. 63. A) archi neu-
rali e costole; B) centri ed elementi indeterminati. Vedi Appendice 1 o
risvolti di copertina per le abbreviazioni.

Fig. 65 - Close-up of the pelvic region of Scipionyx samniticus under
properly grazing light, which highlights the presence of a caudal verte-
bral centrum under the right iliac blade (shadowed relief, and fractures
marked by the arrows). Scale bar = 2 mm. See Appendix 1 or cover
flaps for abbreviations.

Fig. 65 - Particolare della regione pelvica di Scipionyx samniticus in
luce radente, che evidenzia la presenza di un centro vertebrale caudale
al di sotto della lama iliaca destra (rilievo ombreggiato e fratture indi-
cate dalle frecce). Scala metrica = 2 mm. Vedi Appendice 1 o risvolti di
copertina per le abbreviazioni.

against the right ilium and, in its cranial half, is slightly
superimposed to the dorsal edge of the iliac blade. The
craniodorsal quarter of the neural arch is markedly cor-
rugate, possibly representing the articular surface of the
transverse process. Ventral to the arch of the S4, and well-
detached from it, the relief of a quadrangular element can
be seen, under oblique light, under the right iliac blade
(Fig. 65). Both shape and size fit a sacral centrum and,
based on its position, it can be referred to S4. The neural
arch of S5 is also detached from its centrum. Ventral to
the spine, the exposed caudal half is laterally reinforced
by a crest. This crest widens ventrally to form a rugose
scar that represents the articular surface for the 5" sacral
rib. A very similar undescribed crest is visible in the last
sacral of Sinocalliopteryx (Ji et al., 2007a: fig. 4b). The 5%
sacral rib of Scipionyx has moved slightly cranially and,
together with another element of uncertain attribution,
covers the cranial half of the 5*" neural arch. The centrum
of S5 lies in dorsal view, between the right postacetabular
iliac blade and the intestine, exposing a wide neural canal
and both neurocentral articular surfaces. The centrum of
S5 has flat articular surfaces for the adjacent centra.

Caudal vertebrae - Only 9 proximal vertebrae of the
tail of Scipionyx are preserved (Figs. 66-67). Presumably,
the tail was longer than the skull-presacral length (see
Skeletal Reconstruction And...), given that in complete
coleurosaurian skeletons, especially in compsognathids,
the intact tail is definitely longer than the skull-presacral

no: CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 66 - Caudal vertebrae and haemal arches of Scipionyx samniticus. Scale bar = 5 mm.
Fig. 66 - Vertebre caudali e archi emali di Scipionyx samniticus. Scala metrica = 5 mm.

TM caudala
| caudal5
di caudal 6

| indeterminate bone

DM cauda 7

caudal 8

_ caudal 9

RI haemal arch

Fig. 67 - Line drawings of the bones illustrated in Fig. 66. A) neural and haemal arches; B) centra and indeterminate bones.
Fig. 67 - Disegni al tratto delle ossa illustrate in Fig. 66. A) archi neurali ed emali; B) centri e ossa indeterminate.

length, irrespective of the ontogenetic stage (Kobayashi
& Lii, 2003). In the preserved series, no transition point is
seen, as is demonstrated by the persistence of transverse
processes.

The caudal neural spines are markedly low, proximo-
distally elongate and shorter than the transverse process-

es. They bear well-developed zygapophyseal articulations
that overhang their centrum. As in most of the preced-
ing vertebrae, the neural arches are more or less aligned
and articulated, but not fused to their centra, so that the
neurocentral sutures are open and the centra are detached
to varying degrees and variably reoriented. The neural

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 73

arches are mainly preserved in lateral view, but the diage-
netic compression caused a slight rotation that has partly
exposed the left zygapophyses and transverse processes.
This laterodorsal exposure is particularly apparent in Ca3
and Ca$ (Fig. 68).

The caudal centra are long and platycoelous, as in
Compsognathus (Peyer, 2006). From Cal, which lies in
full dorsal view, the centra rotate gradually along the se-
ries up to Ca4, which is in lateral view. This variety of
views allows to reconstruct many structures (see below),
despite the centra of Ca2 and Ca3 being slightly covered
ventrally by patches of soft tissue. The centra of CaS and
Ca6 are no longer aligned along the proximodistal hori-
zontal plane. Their contact has moved so far ventrally
that their adjacent articular faces are perpendicular rather
than parallel. An oblique fracture, interrupting the slab in
which the caudosacral portion of the specimen lies, has
truncated both Ca$, of which the proximal half of the cen-
trum and neural arch are preserved, and Ca9, of which the
centrum is completely missing and only a proximal por-
tion of the neural arch is preserved.

The neural spines of Cal and Ca2 are taller and more
rounded than the successive ones. Notably, the proximal
end of the spine of Cal seems to be in continuity with
the caudal end of the spine of SS (Fig. 69). From Ca3,
the neural spines become lower than long, more elon-
gated proximodistally and tilted backwards. The spines
are tilted backwards also in Compsognathus, but they dif-
fer from those of Scipionyx in being markedly higher and
rectangular in shape (Peyer, 2006). From Ca4, the spines
of Scipionyx develop a slight concavity at mid-length of
the dorsal edge (Fig. 68). However, contrary to Sinosau-
ropteryx (Currie & Chen, 2001) and some other theropods
(e.g., allosauroids), in Scipionyx this concavity does not
really separate the neural spine in a proximal emargination
(accessory neural spine) and a distal spine, at least in the
preserved, proximalmost sequence. Thus, Scipionyx pres-
ents a condition intermediate between Sinosauropteryx
and Compsognathus, in which not even a slight concavity
has been described (Peyer, 2006). Bifid spines are also not
reported in other coelurosaurs such as Ornitholestes and
Coelurus (Carpenter et al., 2005b).

Fig. 68 - Shaded drawing of the 5% caudal vertebra and 5° haemal arch of Scipionyx samniticus. See Appendix 1 or cover flaps for

abbreviations.

Fig. 68 - Disegno ombreggiato della 5° vertebra caudale e del 5° arco emale di Scipionyx samniticus. Vedi Appendice 1 o risvolti di

copertina per le abbreviazioni.

74 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 69 - Close-up of the caudosacral region of Scipionyx samniticus,
showing the seemingly fused neural spines of the 5 sacral and 1" caudal
vertebra. For assr, see also Fig. 79. Scale bar = 1 mm. See Appendix 1
or cover flaps for abbreviations.

Fig. 69 - Particolare della regione caudosacrale di Scipionyx samniticus,
con le spine neurali della 5° vertebra sacrale e della 1° caudale in appa-
rente contatto suturale. Per assr, vedi anche Fig. 79. Scala metrica= 1 mm.
Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

The passage between the spines and the postzygapo-
physes is marked by a margin that appears remarkably
concave in lateral view (right postzygapophyses), whereas
it appears as a ridge subdividing the dorsal surfaces of the
postzygapophyses in dorsal view (left postzygapophyses).
The zygapophyseal articular surfaces seem to be inclined
about 30-45° with respect to the medial sagittal plane.

All the preserved transverse processes are robust and
project perpendicularly to the direction of the vertebral
column. Along the series they also become slightly in-
clined distally and seem to be broader laterally than at their
base, with more concave proximal and distal margins. No
accessory transverse processes, such as those described
by Currie & Chen (2001) in the proximal vertebrae of the
tail of Sinosauropteryx, can be seen in Scipionyx. In Ca2,
the base of the arch, which is not hidden by the transverse
process, distinctly shows its ventral margin. It has a bi-
lobed shape that is more marked than in D12 and D13.

On account of the rotation described above, the floor of
the neural canal and the rugose neurocentral articular sur-
faces of Cal-Ca3 are visible (Fig. 70). The passage from
the neurocentral surfaces to the articular and lateral faces of
the centra is abrupt and angled. This, together with the fact
that the ventral concavity of the centra is very faint, ren-
ders the proximal caudal centra markedly parallelepiped-
like (box-like), as in Compsognathus (Peyer, 2006), Ju-
ravenator (GGhlich et al., 2006), Ornitholestes, Coelurus
(Carpenter et a/., 2005b), deinonychosaurs (Makovicky &
Norell, 2004; Norell & Makovicky, 2004) and basal birds
(e.g., Rauhut, 2003). Besides the ventral concavity, the
centra of Scipionyx ventrally also lack distinct proximal
and distal pedicels for the articulations with the chevrons,
that are well-developed in ornithomimosaurs (Kobayashi
& Lii, 2003; Makovicky er a/., 2004). The centra of Scipio-
nyx are bulky in appearance, with the 1 centrum only 4/3
longer than its width. From Cal, the centra gradually in-

Fig. 70 - Close-up of the proximal caudal vertebrae of Scipionyx sam-
niticus. The floor of the neural canal and the neurocentral articular sur-
faces of the box-like centra are exposed. Left zygapophyses labelled in
black, right zygapophyses in white. Scale bar = 1 mm. See Appendix 1
or cover flaps for abbreviations.

Fig. 70 - Particolare delle vertebre caudali prossimali di Scipionyx sam-
niticus. I centri, di forma squadrata, espongono il pavimento del canale
neurale e le superfici articolari neurocentrali. Zigapofisi sinistre in nero,
zigapofisi destre in bianco. Scala metrica = 1 mm. Vedi Appendice 1 o
risvolti di copertina per le abbreviazioni.

crease in length proximodistally, as occurs in Juravenator
(Géhlich et al., 2006: tab. 1). The last undisturbed centrum
of Scipionyx (Ca7) is about twice as long as tall.

Large pneumatopores are absent, but a single, small
foramen is present at about 1/3 of the proximodistal length
from the proximal margin and at about 1/3 ofthe dorsoven-
tral height from the dorsal margins on the lateral surface
of the centra of Ca4-Caé (Figs. 68, 71). Successive centra
seem to lack this foramen, whereas in the preceding ones
they cannot be observed because the centra are exposed
mainly in dorsal view. The pneumatic or neurovascular
nature of the foramina cannot be ascertained. The absence
of large pneumatopores and the presence of small foram-

Ra

Fig. 71 - Pneumaticity of the caudal vertebrae of Scipionyx samniticus
(arrows point to pneumatopores). Scale bar = 2 mm.

Fig. 71 - Pneumaticità nelle vertebre caudali di Scipionyx samniticus (le
frecce indicano pneumatopori). Scala metrica = 2 mm. Vedi Appendice
1 o risvolti di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 75

ina were found on the middle lateral surfaces of the first
9 caudal centra in 7yrannosaurus (Brochu, 2003). Among
coelurosaurs, no evidence of pneumaticity has been found
in compsognathids, Ornitholestes, Coelurus (Carpenter et
al., 2005b), ornithomimosaurs (Makovicky et al., 2004)
or Falcarius (Kirkland et al., 2005), whereas single pleu-
rocoels have been reported close to the base of the neural
arch in the proximal caudals of some derived coelurosaurs
such as the Oviraptorosauria (Osmélska ef a/., 2004). The
alleged compsognathid Orkoraptor has caudal vertebrae
with a single pair of small pleurocoels on each side (No-
vas et al., 2008a); however, the phylogenetic position of
this fragmentary taxon is not yet clear, it belonging prob-
ably to a group of non-coelurosaurian tetanurans (see
Comments in Phylogenetic Analysis).

Haemal arches - In Scipionyx, two complete haemal
arches (=chevron bones) are exposed (Figs. 66, 67A).
They lie in lateral view in an intercentral position, be-
tween Ca4-CaS and CaS-Ca6. A third one, preserving only
its proximal portion, is located between Ca6 and Ca7. In
theropods, the haemal arches are usually present from the
proximalmost intercentral positions, with a certain degree
of individual variability (e.g., Weishampel et al., 2004).
For example, in Compsognathus, chevrons are present
from the first intercentral position (i.e., between Cal and
Ca2) (Peyer, 2006), whereas in Huaxiagnathus they are
reported to appear first between Ca3 and Ca4 (Hwang ef
al., 2004). In Scipionyx, the presence/absence of haemal
arches in the first three intercentral positions cannot be
ascertained because of the presence of a large mass of
soft tissue, including muscular remains, which covers the
ventralmost portion of the caudal skeleton. The thin bone
portion aligned to the ventral margin of the centrum of
Ca4 may be part of the shaft of a haemal arch, whereas the
outline of an element matching in size and orientation the
other haemal arches can be seen with CT scanning at the
Ca2-Ca3 intercentral position.

The chevrons of Scipionyx are generally slender and
feebly recurved, like those of all neotetanurans (Holtz et
al., 2004) except the most derived Maniraptora. Proxi-
mally, they bear cranial and caudal articular facets that,
when articulated, extend beneath the ventral surfaces of
the adjacent caudal centra (Figs. 68, 72). In the two fully
preserved elements, the cranial articular facet appears
weakly convex and slightly more extended than the cau-
dal one, which is markedly concave. The portion of bone
between the two facets forms an angle greater than 100°.
A pointed, cranial process that is little differentiated from
the shaft is also present (Figs. 68, 72): it is comparable
in size to that of Sinocalliopteryx (Ji et al., 2007a) and
Juravenator (Gòhlich et al., 2006), but markedly smaller
than in A//osaurus (Madsen, 1976) and Majungasaurus
(O’Connor, 2007), where it approaches the size of the
articular facets. Moreover, in these taxa the angle is less
than 90° and the orientation of the facets is symmetric
with respect to the proximal half of the shafît.

In Scipionyx, the marked concavity of the caudal ar-
ticular facet indicates an intimate connection with the cra-
nial margin of the distal vertebral centrum. This character,
together with the asymmetric position of the facets with
respect to the shaft, suggests that în vivo the proximal por-
tion of the shaft of the proximal haemal arches diverged
less than 40° from the long axis of the tail. The divergence

Fig. 72 - Close-up of the 5 haemal arch of Scipionyx samniticus. Note
the distal flattening and fusion of the two rami. Scale bar = 1 mm. See
Appendix 1 or cover flaps for abbreviations.

Fig. 72 - Particolare del 5° arco emale di Scipionyx samniticus. Notare
l’appiattimento distale e la fusione dei due rami. Scala metrica = 1 mm.
Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

is of about 30° in Sinocalliopteryx (Ji et al., 2007a). The
curvature of the shaft, uniform for almost 2/3 the length of
the haemal arch, disappears distally, where the two coun-
terlateral rami meet and become a single mediolaterally
flattened element.

The chevrons of the other compsognathids are more
or less similar in general shape. Like in Scipionyx, those
occupying the intercentral positions 4-5 and 5-6 seem
to be slender, rod-like and possibly flattened distally in
Huaxiagnathus (Hwang et al., 2004: fig. 4C), in the Ger-
man Compsognathus (Ostrom, 1978) and in Juravenator
(G6hlich et al., 2006: pl. 7, fig. 3). They appear compara-
tively broader and more recurved in the French Compso-

76 CRISTIANO DAL SASSO & SIMONE MAGANUCO

gnathus (Peyer, 2006), and even broader, gently curved
and remarkably spatulate in Sinosauropteryx (Currie &
Chen, 2001; Ji et a/., 2007b). Given the apparent variabil-
ity of the chevrons from neighbouring intercentral posi-
tions in the above-cited complete tails, these differences
alone do not deserve much attention.

Ribs

Scipionyx possesses 21, possibly 22, pairs of straight
to moderately curved presacral ribs (Figs. 49, 55). Those
of the right side are all visible, at least for part of their

length, whereas the left ones are not easily identifiable,
emerging only for short tracts, especially in the cervical
region. All cervical and cranial dorsal ribs are clearly di-
chocephalous; caudal dorsal ribs maintain long capitula,
but the tubercula shorten gradually.

Cervical ribs - Nine pairs of cervical ribs are present,
from the axis on (Fig. 73). The atlas does not bear ribs, as
is the case in all neotheropods (Weishampel et a/., 2004).
The cervical ribs of Scipionyx are more robust proximally
and much more elongate distally than previously thought
(Dal Sasso & Signore, 1998a); however, they remain
thicker than in other compsognathids, such as Juravena-

Li cervical 2

| cervical3

| Vos

I i cervical 4

cervical 6
cervical 7

cervical 8

i cervical 5 i cervical 9

i cervical 10

| indeterminate bone

calcite vein

Fig. 73 - Line drawings of the cervical ribs of Scipionyx samniticus, illustrated in Fig. 49. A) right ribs; B) left ribs and indeterminate

elements.

Fig. 73 - Scipionyx samniticus. Disegno al tratto delle costole cervicali illustrate in Fig. 49. A) costole destre; B) costole sinistre ed

elementi indeterminati.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY sf,

tor (Gòhlich & Chiappe, 2006: fig. 1b). With the excep-
tion of Crl0, all the right cervical ribs have capitula still
in place and articulated to their vertebral centra, thus, hid-
ing all parapophyses. On the other hand, all the tubercula
except those of Cr3 and Cr4 have lost their contact with
the diapophyses.

The right axial cervical rib is characterised by a small
head, with adjacent, faintly diverging tuberculum and ca-
pitulum. Therefore, as already written, it can be consid-
ered dichocephalous, a condition shared by the tetanuran
theropods (Holtz et al., 2004). At a first glance, the shafts
(stilyform processes sensu O°Connor, 2007) of the axial
ribs seem to be as long as the axial centrum. However,
observing the fossiliferous slab at a high magnification
and inclined at 45°, it is apparent that there is continuity
between them and the tiny bones ventral to the centrum
of C3. Thus, despite their delicate nature, both right and
left Cr2 are preserved for their entire length, parallel and
paired, and are as long as two vertebral centra (Fig. 74).

The head of Cr3 is the most fragmented of the right
side. Its seemingly small size is due to the fact that only
the capitulum is still connected to the shaft. The tuber-
culum is, in fact, detached but still articulated to the di-
apophysis. The proximal portion, which runs ventral to
the centrum of C3 before disappearing under the head of
the right Cr4, is thicker than that of Cr2. Observing the
order of superimposition of the right and left shafts, and
following a process of elimination, the distal end of the
right Cr3 can be seen passing under the head of the Cr6
and terminating juxtaposed to the left ceratobranchial I.
Thus, from Cr3 the cervical ribs already reach the length
of three vertebral centra. As for Cr2, also the distal end of
the left Cr3 is well-preserved and lies close to the distal
end of the right Cr3, both forming a V and both being jux-
taposed to the left ceratobranchial I (Fig. 54).

In the right Cr4, tuberculum and capitulum clearly di-
verge and the transition between the head and the shaft
is more marked, not only by an abrupt decrease in size
in transverse section, but also because the shaft seems to
change direction, paralleling the concavity of the ventral
margin of the centrum.

An abrupt directional change not related to fracturing
is present also in the right Cr5, between head and shafît. In
our opinion, this pattern is not completely interpretable as
an artefact of preservation (see Skeletal Taphonomy) but

Fig. 74 - Photograph taken in grazing (ventrolateral) view of the neck
of Scipionyx samniticus, showing origin and end of the 2°° cervical ribs
(black arrows), and their continuity under the 2"° and 3"° vertebral centra
(red arrow). See Appendix 1 or cover flaps for abbreviations.

Fig. 74 - Vista radente (ventrolaterale) del collo di Scipionyx samniticus,
che mostra origine e terminazione delle seconde costole cervicali (frecce
nere), e la loro continuità sotto il 2° e 3° centro vertebrale (freccia rossa).
Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

reflects, in part, the natural arrangement. Curvature of the
neck of the rib is, in fact, preserved also in the right Cr7
and, symmetrically, in both right and left Cr8, the shafts
of which, unlike the preceding ones, are not close to the
centra of their respective vertebrae. Like Cr3, we identi-
fied the end of the left Cr5 by looking at the specimen
inclined at 45°. It lies at the level of the caudal margin of
the centrum of C7, indicating that Cr$ is as long as three
centra as well.

The right Cré lacks the tuberculum and part of the
neck; it disappears in a caudal direction under the cen-
trum of its corresponding vertebra. Two short segments
of uncertain attribution are present between the right Cr6
and the ventral surface of the centrum.

The pair of Cr7 represents the longest cervical ribs.
They lie ventral to cervical centra 7, 8 and 9, which are
the most elongate craniocaudally. They are also the most
clearly exposed ribs, as their mid portion is not hidden
by the successive centrum. They preserve a natural dorsal
outline with a marked neck convexity, a faint concavity in
the mid portion of the shaft, and a distal portion which is
almost straight.

The right Cr7 and Cr8 are the cervical ribs which pos-
sess the dorsoventrally most expanded heads; they have
a markedly elongate tuberculum projecting towards an
almost horizontal diapophysis, and an apparently shorter
capitulum. The two processes are connected via a bony
lamina, the capitotubercular web (sensu O’ Connor, 2007).
The concavity of the tuberculum of the right Cr$8 is the
best preserved of the cervical series.

The right Cr9 is fractured but well-exposed all along
its length; it is shorter than the preceding ones, as it does
not reach the length of three vertebral centra, which, in
addition, are slightly shorter here than the preceding ones.
The neck of the rib displays the above mentioned natural
caudodorsal curvature; however, in the mid portion the
dorsal concavity is accentuated by the ventral displace-
ment of the centrum of C10, which has caused the fracture
of the distal portion of the shafît.

The heads of the left Cr9 and Cr10 emerge consider-
ably in the space between the neural arches and the centra
of C9 and C10. The almost erect tubercula are buttressed
ventrally by the capitotubercular web (Fig. 75); also, part
of the shaft of the left Cr9 can be seen directed caudally.
The right Cr10 lacks the shaft but preserves a markedly
enlarged head, which is comparable in size to the first dor-
sal ribs but maintains the shape of the cervical ones. In
fact, differently to the dorsal ribs (see below), the tubercu-
lum projects dorsally, where it contacted the diapophysis
(they became separated by a calcite vein — see Skeletal
Taphonomy), whereas the capitulum is almost aligned
with the major axis of the shaft. Thus, the right Cr10 pres-
ents a transitional morphology and, similar to its vertebra,
can be considered a cervicodorsal rib. Unfortunately, both
shafts of Crl0 are missing, so we cannot compare their
length with that of the first dorsal rib.

The cervical ribs of Scipionyx resemble those of the
other compsognathids in being elongate and very thin.
Like in Scipionyx, the cervical ribs of Compsognathus are
reported to be very delicate and slightly curved, at least
slightly longer than the cervical vertebrae in the French
specimen (Peyer, 2006: figs. 2B, 6), and more than double
the length of the cervical centra in the German specimen
(Ostrom, 1978). The cervical ribs of Sinosauropteryx are

78 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 75 - Close-up of the head of the 9° right cervical rib of Scipio-
nyx samniticus. Scale bar = 0.5 mm. See Appendix 1 or cover flaps for
abbreviations.

Fig. 75 - Scipionyx samniticus, particolare della testa della 9° costola
cervicale destra. Scala metrica = 0,5 mm. Vedi Appendice 1 o risvolti di
copertina per le abbreviazioni.

reported to be very thin and not as long as those reported
for Compsognathus (Currie & Chen, 2001), given that
they are slightly longer than one cervical centrum (Ji et
al., 2007b); based on figures, however, they seem to be as
long as two centra, or even more in some cases, just like
in Scipionyx and Compsognathus. Slender cervical ribs,
longer than the centra to which they attach, are reported
also in Huaxiagnathus (Hwang et al., 2004) and Sinocal-
liopteryx. In the latter, they are thread-like and extremely
long, extending at least along three cervical centra (Ji et
al., 2007a: fig. 4a). The cervical ribs are rather short in
Maniraptoriformes, whereas longer cervical ribs can be
found among non-coelurosaurian theropods: in A//osau-
rus, they overlap at least one-half of the following cen-
trum (Madsen, 1976), and in Coelophysoidea and Neo-
ceratosauria, they surpass caudally at least three vertebrae
(up to six in some species) beyond their origin (Tykoski
& Rowe, 2004).

Dorsal ribs - The dorsal ribs of Scipionyx consist of
12, possibly 13, pairs of gently curved, craniocaudally
compressed rods with tubercula shorter than capitula
and median longitudinal grooves (costal grooves) em-
phasised by diagenetic compression (Fig. 76A). There
are neither uncinate processes, nor ossified ventral ele-
ments, as the so called “abdominal ribs” figured by Dal
Sasso & Signore (1998a: fig. 2) are actually lateral gas-
tralia.

It is not easy to distinguish between the right and left
ribs, especially at their distal ends, because diagenetic
compression has left them often on the same plane, with
the latter sometimes pushed onto the former. However,
following the right ribs from the heads, and observing the
differences in height of the shaft fragments, it is possible
to outline each right rib and to estimate which fragments
belong to the left side (Fig. 76B). It is then apparent that
almost every left rib terminates cranially to the corre-
sponding right rib. Some supernumerary elements have
been also identified in the chest of Scipionyx: they repre-
sent allochthonous bones (see Gastric Contents).

Right dorsal ribs 1 to 5 lie almost undisturbed and al-
most parallel to each other along their whole length; Dr6
and Dr7 are the most fractured, whereas right Dr8 and
Dr12 seem to be complete and still showing their curva-
ture. Dr 9-11 are uplifted and slightly rotated clockwise,
but still entire. As mentioned, a small fragment parallel-
ing the shaft of the right Dr12 and preserved against the
centra of D13 and SI is tentatively interpreted as belong-
ing to a Drl3 (Fig. 76A). Currie & Chen (2001) reported
that two specimens of Sinosauropteryx have 13 dorsal
vertebrae and 13 pairs of dorsal ribs; according to Peyer
(2006), a rib was certainly present on the last dorsal ver-
tebra of the French Compsognathus, although it was not
preserved in the fossil specimen. In the Ornithomosauria
and in the Dromaeosauridae, a small but single-headed
rib articulates with the last dorsal vertebra (Makovicky
et al., 2004; Norell & Makovicky, 2004). A boomerang-
shaped bone was tentatively identified by Brochu (2003)
as a vestigial rib of the last dorsal vertebra in the giant
coelurosaur 7yrannosaurus, whereas in the neoceratosaur
Majungasaurus (O’Connor, 2007), the Dr12 is already
vestigial and D13 lacks one rib altogether.

Following the curvature of the shafts, the shape of the
trunk of Scipionyx can be reconstructed: the cranial dorsal
ribs have an almost straight median tract, indicating that
the thoracic region was relatively flat laterally; the caudal
dorsal ribs are more evenly curved all along their shafts
and, thus, delimited a more rounded abdominal region
(see Skeletal Reconstruction And...).

Similar to the last cervical rib, the first dorsal rib
can be considered transitional. The proximal portion is
markedly more expanded than the shaft, as in Crl0; the
shaft is also considerably shorter than those of all the
other dorsal ribs of the thoracic region and is compa-
rable in length to that of Cr9. Moreover, this rib can be
considered to be the first dorsal rib because its tubercu-
lum and capitulum are almost the same size as those of
the right Dr2 and, above all, because the tuberculum is
almost aligned with the long axis of the shaft whereas
the capitulum projects ventrally towards the parapophy-
sis, contrary to what is observed in the right Crl0 but
like the situation in the cranial dorsals. This morphology
is slightly emphasised by the fact that the capitulum of
Drl was separated from the head during diagenesis, be-
cause of the formation of a calcite vein.

The heads and necks of Dr2-Dr4 are particularly robust,
with tubercula that are deeper than capitula, and broad and
thin capitotubercular webs (Fig. 77). The shafts lack the
peculiar craniocaudally expanded proximal portion vis-
ible in the Ornithomimosauria (Makovicky et al., 2004),
and gradually taper all along their length. The shafts of
the left and right Dr2 are still relatively slender and short
with respect to the successive ones. The shafts of Dr3-Dr7
are long and more or less constant in length, Dr6 and Dr7
being the longest, as in Compsognathus (Peyer, 2006).
According to Hwang et al. (2004), in Huaxiagnathus the
longest pairs of ribs are attached to the 4" to 8! dorsal
vertebrae. In Scipionyx, the shafts decrease gradually in
length from Dr8 to Dr12. The tubercula shorten between
Dr$ and Dr7; they are poorly visible because the shaft of
the preceding rib covers exactly the head and the neck of
the successive one. This shortening is so considerable that
it becomes difficult to locate the tubercula from Dr8 on
because the curvature of the shaft is almost uniform up to

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 79

Di dorsal 1 dorsal 6
Di dorsal 2 dorsal 7
È dorsal 3 dorsal 8
si dorsal 4 dorsal 9
E dorsal 5 Li dorsal 10

dorsal 11

Sd dorsal 12
23 dorsal 13 |
CI indeterminate bone |

calcite vein

Fig. 76 - Line drawings of the dorsal ribs of Scipionyx samniticus, illustrated in Fig. 55. A) right ribs; B) left ribs and indeterminate

elements.

Fig. 76 - Scipionyx samniticus. Disegno al tratto delle costole dorsali illustrate in Fig. 55. A) costole destre; B) costole sinistre ed

elementi indeterminati.

the capitulum (Fig. 76A). In contrast, the capitula remain
almost the same length in all dorsal ribs.

Dr3 and Dr4 are the most robust elements of the se-
ries. The distal end of the ones of the left side are well-
exposed and intact. Rather than tapering to a point, these
ribs terminate in an expanded, cup-like surface (Figs.
76B, 78) that articulated with the sternal complex, which
is not preserved in this specimen of Scipionyx (see also
Ontogenetic Assessment), through a system of sternal ribs
and/or costal cartilages (see Pectoral Girdle; Godfrey &
Currie, 2004; and references therein). Similar cup-like

depressions, interpreted as connections with sternal ele-
ments, are reported also by Currie & Chen (2001) in the
first two pairs of dorsal ribs of Sinosauropteryx. One very
similar expansion is likely preserved at the distal end of
the left Dr3 in Juravenator (Dal Sasso & Maganuco, pers.
obs., 2006 on unpublished photographs by Gohlich &
Chiappe), and two expanded articular surfaces seem to be
present at the distal end of the first pair of elongate dorsal
ribs (?Dr3) in Sinocalliopteryx (Ji et al., 2007a: fig. 1). In
Majungasaurus (O°Connor, 2007), a similar morphology
is present in Dr2 and Dr3, which terminate with blunt,

80 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 77 - Close-up of the heads of the 2°° and 3" right dorsal ribs of
Scipionyx samniticus. Scale bar = 1 mm. See Appendix 1 or cover flaps
for abbreviations.

Fig. 77 - Particolare delle teste della 2° e 3° costola dorsale destra
di Scipionyx samniticus. Scala metrica = 1 mm. Vedi Appendice 1 o
risvolti di copertina per le abbreviazioni.

Fig. 78 - Close-up of the cup-like distal end (arrow) of the 3" left dorsal
rib of Scipionyx samniticus. Scale bar = 0.5 mm. See Appendix 1 or
cover flaps for abbreviations.

Fig. 78 - Scipionyx samniticus. Particolare della terminazione distale a
coppa (freccia) della 3° costola dorsale sinistra. Scala metrica = 0,5 mm.
Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

squared-off ends. Squared-off distal ends, possibly for ar-
ticulation with the cartilaginous sternum or sternal ribs,
are also reported by Kobayashi & Lii (2003) in Dr2-Dr8
of Sinornithomimus.

Sacral ribs - Two, or possibly 3, disarticulated bones
in the sacral region of Scipionyx can be identified as sacral
ribs (Figs. 63-64). The first element is enclosed between
the caudal margin of the ilium and the first caudal verte-
bra. Although fossilised caudally to Sr5 (see below), we
tentatively consider it to be the Sr4, based on the strong
resemblance both in shape and in relative size to the one
in A//osaurus (Madsen, 1976). If this hypothesis is right,
it represents the right Sr4 exposed in lateral view and ro-
tated 180°, as shown by the sutural scar for the ilium vis-
ible on its more robust end (Fig. 79). A second element,
which is certainly a sacral rib, covers the cranial third of
the neural spine of SS. Again, based on the strong resem-
blance with the shape of the 5° sacral rib of A//osaurus

mortes

Fig. 79 - Close-up of the disarticulated sacral ribs of Scipionyx sam-
niticus. Scale bar = 2 mm. See Appendix 1 or cover flaps for abbrevia-
tions.

Fig. 79 - Le costole sacrali disarticolate di Scipionyx samniticus. Scala
metrica = 2 mm. Vedi Appendice 1 o risvolti di copertina per le abbre-
viazioni.

(Madsen, 1976: pl. 27F), we interpret the exposed surface
as the lateral side of the right SrS. Larger than Sr4, Srs
has a rhomboidal shape and bears a caudal ridge that is
expanded ventrally to form a robust, craniocaudally en-
larged base for contact with the ilium (Fig. 79). This ele-
ment was misinterpreted as the neural spine of SS by Dal
Sasso & Signore (1998a). The difference in size between
Sr4 and SrS (the latter is larger than the former) can be ex-
plained by the fact that the iliac blades, when seen in dor-
sal/ventral views, probably diverged at both cranial and
caudal ends, increasing the distance between the sacral
vertebral centra and the blades themselves.

Athird possible sacral rib lies just underneath Sr$, emerg-
ing a little dorsal to it: identified originally as a sacral rib by
Dal Sasso & Signore (1998a), it seems to be the mirror im-
age of SrS and may indeed represent the left Sr5. However,
as it is mostly hidden and does not present unequivocal fea-
tures, we prefer to regard it as an indeterminate element.

Gastralia

Exceptionally preserved in situ, there are at least
18 pairs of delicate, rod-like medial gastralial elements
(Figs. 80-81). The medial ends lie still aligned to form the
ventral margin of the thoracoabdominal region, in ideal
continuation with the pubic foot. Their imbricate arrange-
ment (sensu Claessens, 2004) is much better maintained
in the rows immediately caudal to the left wrist, where a
right medial element still lies almost in contact with two

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 81

Fig. 80 - Gastralia of Scipionyx samniticus. Scale bar = 5 mm.

Fig. 80 - Gastralia di Scipionyx samniticus. Scala metrica = 5 mm.

Fig. 81 - Line drawing of the bones illustrated in Fig. 80. The red outline indicates a possibly aberrant element. See Appendix 1 or
cover flaps for abbreviations.
Fig. 81 - Disegno al tratto delle ossa illustrate in Fig. 80. Il contorno rosso indica un probabile elemento aberrante. Vedi Appendice 1
o risvolti di copertina per le abbreviazioni.

medial elements from the left side of the body, one at the
medial end and the other farther down the shafît, in a zigzag
fashion. The successive elements, although still in place,
are not preserved in articulation. Generally, the left ele-
ments expose their dorsal surface, whereas the right ones
are visible in ventral view. Both spacing and length of the
medial gastralia decrease considerably towards the pubis.
At least S slightly longer lateral gastralia are preserved
in association with some of the mid medial elements, to
which they articulate in parallel, overlapping the intestine
of Scipionyx in a dorsocaudal direction.

The cranialmost element, thicker than the mid gastralia
but much more slender than the dorsal ribs, is incomplete
and partly hidden by the right humerus and, maybe (see
below), by the right radius. The most expanded portion,
emerging from a plane under the humerus, terminates at
a distance equal to 1/3 the length of the whole preserved
portion. Here, the shaft curves and forms an obtuse angle
of about 160° before tapering abruptly, becoming as thin
as the medial portion of the successive gastralia. This thin
portion seems to continue in a caudal direction, preserved
as an imprint in the matrix. Following this imprint, another
thin bone fragment can be seen close to the right radius,
immediately covered by the successive gastralium and,
possibly, by the radius itself (Fig. 81). The whole element
was, therefore, at least twice as long as the preserved por-
tion. Based on shape and size, we interpret this element as
one arm of the unpaired cranial chevron-shaped gastralium
(sensu Claessens, 2004). The first regular row of gastralia is
composed of only two straight medial elements in Sinocal-

liopteryx (Ji et al., 2007a) and Juravenator (G6hlich et al.,
2006). In the latter, the two elements clearly form a cranial
chevron-shaped gastralium, as they are fused in a single
element that is twice as thick as the successive ones.

In Scipionyx, caudal to the chevron-shaped gastralium,
3 rows of medial gastralia run caudodorsally. As they are
sandwiched between the forearms, they cannot be de-
scribed in detail. The medial ends of the successive 6 rows
are hooked and considerably expanded (Fig. 82). Accord-
ing to Claessens (2004), in small theropods the wing-like
expansion of the medioventral articular facet is not as pro-
nounced as in larger theropods, and may be absent. In Sci-
pionyx, however, just caudal to the point of abrupt curva-
ture, there is a point of maximum expansion which cor-
responds to the wing-like expansion of the medioventral
facet (mvf). A thin continuation, directed medially and
terminating in a pointed end, bears the mediodorsal facet
(mdf). In these mid gastralia, the largest transverse diam-
eter is at mid-length; from this point on, in lateral direction,
the medial gastralia flatten and taper, forming a long sutural
surface for the contact with the lateral gastralia (Fig. 83).

The medial portion of the lateral gastralia is as thin
and flat as the complementary lateral portion of the medial
ones. Beyond the sutural surface, which is oriented medio-
caudally, their shafts become rounded in cross section and
their diameters become considerably smaller, measuring
almost half the diameter of the medial elements to which
they articulate. On account of the small size of the indi-
vidual, the lateral gastralia of Scipionyx have a hair-like
appearance (Figs. 80-81). The lateral gastralia are very

82 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 82 - Cranial medial gastralia of Scipionyx samniticus. Close-up of
the medial ends. Scale bar = 1 mm. See Appendix 1 or cover flaps for
abbreviations.

Fig. 82 - Gastralia craniali mediali di Scipionyx samniticus. Partico-
lare delle estremità mediali. Scala metrica = 1 mm. Vedi Appendice 1 o
risvolti di copertina per le abbreviazioni.

fragmented and lean against the intestine for about 2/3
of their length. The distalmost fragments reach the dorsal
margin of the descending loop of the duodenum. Mea-
sured at this level, the lateral gastralia are slightly longer
than (about 8/7) the medial ones.

The caudal medial gastralia consist of some 8 rows of
shorter rods, not preserved in articulation. Cranial and ven-
tral to the ascending loop of the duodenum, they have an
unusual hammer-like conformation: 6-8 bifid medial ends
can be seen (Fig. 84), referable to at least four rows of gas-
tralia. These bifid ends possibly derive from a greater sepa-
ration between the mvf and the mdf. In particular, the end
directed towards the intestine would correspond to the mvf
of the preceding, enlarged and more elongated gastralium;

the shorter and globular opposite end would correspond to
the mdf. The shaft originates at about mid-length of the por-
tion which bears the facets. This conformation possibly in-
creased the length of the sliding surface ofthe mvfin which
the mdf of the successive element was accommodated. As
hypothesised by Claessens (2004), inward sliding of the
mediodorsal joints on gastralial retraction, combined with
outward sliding of the joints upon gastralial protraction,
would have accommodated some of the lengthening and
shortening of the whole gastralial system during protrac-
tion and retraction. According to the model proposed here-
in (Fig. 85), the range of movement (protraction/retraction)
of the gastralia on a plane, described by Claessens (2004) in
other theropods, would result similar in Scipionyx.

The thin fragments of at least 3 lateral gastralia, in di-
rectional continuity with the same number of medial ele-
ments, are preserved on the ascending loop of the duode-
num and appear as long as the most complete preceding
lateral gastralia. In this area too, the lateral gastralia are
shorter than their corresponding medial gastralia.

The caudalmost medial gastralia, which are composed
of 4 rows, are even smaller and overlap each other. They do
not seem to be bifid; however, the better exposed elements
appear to maintain the hooked curvature at the medial end
which characterises the whole series. In any case, unlike
Shenzhousaurus (Ji et al., 2003), the last medial gastralia
of Scipionyx do not form any large caudal chevron-shaped
gastralium (sensu Claessens, 2004). At this level, the lateral
gastralia seem to be absent, or alternatively, they might be
so short and perfectly aligned to the medial ones as to seem
absent. Makovicky et a/. (2004) reported that in ornithomi-
mosaurs the last gastalia are composed of the medial ele-
ments only, which wrap around the cranial expansion of the
pubic foot (absent in Scipionyx). In Sphenodon and in ex-
tant crocodylians, the gastralia are attached to the pubis and

Fig. 83 - Cranial gastralia of Scipionyx samniticus. Close-up of the sutural surfaces between medial and lateral elements. Scale bar
=1 mm. See Appendix 1 or cover flaps for abbreviations.
Fig. 83 - Gastralia craniali di Scipionyx samniticus. Particolare delle superfici suturali tra elementi mediali e laterali. Scala metrica
= 1 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 83

Fig. 84 - Caudal medial gastralia of Scipionyx samniticus. Close-up of
the medial ends and part of the shafts. Scale bar = 1 mm. See Appendix
1 or cover flaps for abbreviations.

Fig. 84 - Gastralia caudali mediali di Scipionyx samniticus. Particolare
delle estremità mediali e di parte delle aste. Scala metrica = 1 mm. Vedi
Appendice 1 o risvolti di copertina per le abbreviazioni.

sternum through midventral ligaments (Claessens, 2004).
In theropods, the attachment of the gastralia to the pubic
bones and sternum was probably similar, and they were
likely embedded in the superficial layers of the M. rectus
abdominis in a similar way (Claessens, 2004). Dal Sasso

/N
/N

& Signore (1998a) reported that in Scipionyx the gastralian
basket seems to constitute effective support for the intestine,
whereas the pubic bones are alien to this function. If such
supporting function was real, it would have been indirect. In
fact, the gastralia are dermal ossifications that cannot attach
directly to the intestine. The latter is loose but suspended by
the mesentery in the peritoneal cavity, which in turn is sur-
rounded by the body wall consisting of different layers of
connective tissue and muscles in which the gastralia are em-
bedded; the whole is surrounded by the dermis. Connective
tissue, muscles and dermis constitute the body wall which
supports the intestines in the abdominal cavity. Therefore,
the intimate connection between the gastralia and the in-
testine, as we see it in the Scipionyx fossil, took place post
mortem, after decomposition of some tissue.

A triangular bony lamina delimited by two gastralial
shafts is related to the duodenum in a similar manner (Fig.
86). The dorsal shaft is superimposed onto the duodenum,
whereas the more robust ventral one makes a U-turn just
opposite to the duodenum itself and then continues in a
cranial direction to contact the bifid medial end of a suc-
cessive gastralium. The function — if any — of such a struc-
ture is unclear, but the fact that the two shafîs are fused to
the lamina and seem to be the result of the splitting of the
shaft of a single medial gastralium suggests that it repre-
sents an aberrant structure. To our knowledge, no similar
structure has been described to date in the literature, at
least in theropod dinosaurs, although different kinds of
aberrant fusion and alteration of the rows have been docu-

Pa
Z-N

Fig. 85 - Schematic modelling of the retraction and protraction movements of the mid-cranial gastralia (top) and mid-caudal gastralia
(bottom) of Scipionyx samniticus as seen in dorsal view (left elements in red, right elements in green), showing the degree of widening
of the lateral abdominal wall (grey area) upon a rotation (A to B) of about 10 degrees. The inward and outward sliding of the joints
during protraction and retraction is appreciable in the mid-caudal gastralia.

Fig. 85 - Scipionyx samniticus. Modello di retrazione e protrazione dei gastralia mediali (in alto, gastralia cranio-mediali; in basso,
gastralia medio-caudali) in norma dorsale, che mostra il grado di espansione della parete addominale laterale (area grigia) in seguito
ad una rotazione di 10 gradi circa (da A a B). In rosso gli elementi del lato sinistro, in verde quelli del lato destro. Lo scorrimento delle
giunture verso l’interno e verso l’esterno, durante la protrazione e la retrazione, è più apprezzabile nei gastralia medio-caudali.

84 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 86 - Close-up of a possibly aberrant gastralial structure of Scipionyx samniticus (white outline). Scale bar = 1 mm. See Appendix
1 or cover flaps for abbreviations.
Fig. 86 - Una probabile struttura gastrale aberrante di Scipionyx samniticus (contorno bianco). Scala metrica =1 mm. Vedi Appendice
1 o risvolti di copertina per le abbreviazioni.

mented (e.g., the bifid element observed in a velocirapto-
rine deinonychosaur by Norell & Makovicky [1997]).
According to Claessens (2004), the variation in the num-
ber of rows between different theropod groups seems to be
related to size, with smaller theropods generally having the
fewest rows of gastralia, and larger theropods appearing
to have the most. Scipionyx, with its 18 pairs of gastralia,
seems to contradict this statement and possess one of the
highest counts among the small theropods. A similar, but
slightly lower, count is found in Compsognathus, which
has about 30-34 elements arranged in 15-17 rows (Peyer,
2006: fig. 2), and in Sinosauropteryx (Ji et al., 2007b), in
which about 15 rows of gastralia are present. Only 12 rows
are present in the complete basket of Sinocalliopteryx (Ji
et al., 2007a), the largest of the compsognathids. At least
14 rows are present in the ornithomimosaur Sinornithomi-
mus (Kobayashi & Lii, 2003). The highest count in small to
medium sized theropods is found in another ornithomimo-
saur, Ornithomimus: a specimen described by Makovicky
(1997) possesses 19 rows of gastralia. The same number
is reported in a subadult specimen of A/bertosaurus and
is possibly exceeded by some large theropods (Claessens,
2004), even if in a large specimen of 7yrannosaurus only
14 rows are described with certainty (Brochu, 2003). 12-
15 rows are present in deinonychosaurs (Russell & Dong,

1993; Norell & Makovicky, 1997), whereas 12-13 are pre-
served in Archaeopteryx (Elzanowski, 2002).

Claessens (2004) observed that in large theropods
(e.g., allosaurids or tyrannosaurids), the lateral gastralia
are much smaller than the medial gastralia, both in length
and diameter, whereas in prosauropods and small thero-
pods (e.g., coelophysids, troodontids, oviraptorids and
dromaeosaurids), the lateral gastralia are 1.5-2.5 times
as long as their medial counterparts. To this list of small
theropods having lateral elements that are considerably
longer than the medial ones, further taxa can be added,
such as troodontids (Russell & Dong, 1993; Currie &
Dong, 2001; Ji et a/., 2005) and compsognathids (Holtz
et al., 2004). In some compsognathids, such as Scipionyx
and Compsognathus (Peyer, 2006: fig. 3a), the lateral gas-
tralia are indeed longer than the medial ones. In contrast,
in Huaxiagnathus (Hwang et al., 2004), Sinosauropteryx
(Currie & Chen, 2001; Ji et a/., 2007b) and Sinocalliop-
teryx (Ji et al., 2007a), the medial gastralial segments are
distincetly longer than the lateral segments. This occurs
also in most ornithomimosaurs (Makovicky e? al., 2004;
Kobayashi & Barsbold, 2005), even though it must be
noted that medial and lateral elements are approximately
equal in length in the most complete row of the basal or-
nithomimosaur Shenzhousaurus (Ji et al., 2003).

APPENDICULAR SKELETON

Pectoral girdle

The pectoral girdle of Scipionyx is preserved in rela-
tive anatomical connection, with its elements forming
a flattened, but still continuous, arc from the left to the
right side (Figs. 87-88). The preservation of the furcula is
remarkable. The furcula is an unpaired median bone de-
rived from the midline fusion of the paired clavicles (e.g.,
Nesbitt ef a/., 2009). Unique to theropods, it is crucial for
understanding the link between dinosaurs and birds.

Scipionyx is not only one of the few dinosaur skeletons
that preserve a furcula, but also one of the even rarer spec-
imens in which the bone is positioned almost like it was în
vivo. Moreover, the simultaneous preservation of the gas-
tralia in a definitely more caudal position and the presence
of a clear hypocleideum in the furcula permit to state that
the two elements, often misinterpreted in past literature,
actually do not have any significant resemblance.

Scipionyx lacks sternal plates. The formerly sup-
posed sternum (“st” in Dal Sasso & Signore, 1998a:

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 85

fig. 2) is actually the proximal portion of the left hu-
merus with its deltopectoral crest. Sternal plates are not
expected to be calcified in a young animal (Gishlick,
pers. comm., 2000). Moreover, the only ossified (and
thus fossilised) theropod sternum so far known out-
side the Maniraptoriformes (e.g., Pérez-Moreno et al.,
1994; Godfrey & Currie, 2004) is doubtful: reported by
Lambe (1917) to be present in A/bertosaurus libratus,
this bone was later re-identified by Claessens (2004)
as the apex of two additional fused cranial chevron-
shaped gastralial rows. Therefore, a cartilaginous ster-
num would be expected also in an adult individual
of Scipionyx (see Ontogenetic Assessment). As men-
tioned, indirect evidence of the sternum is given by the
presence of expanded ends in Dr3 and Dr4 (Figs. 76B,
78), which would have contacted the sternum directly
or via cartilaginous sternal ribs, as is the case in other
non-maniraptoran theropods.

The scapula of Scipionyx, with its dorsalmost mar-
gin incomplete, is only slightly shorter than the humer-
us. With a preserved scapula/humerus ratio of 0.90, it
compares well with both specimens of Compsognathus
(Ostrom, 1978; Peyer, 2006) and with Huaxiagnathus
(Hwang et al., 2004), in which the maximal length of
the scapula approaches that of the humerus. The hu-
meri are comparatively very short in Sinosauropteryx
(Currie & Chen, 2001) and Juravenator (Gòhlich &
Chiappe, 2006), in which the scapula/humerus ratio is
about 1.50 to 1.60. Sinocalliopteryx (Ji et al., 2007a)
presents an intermediate condition (about 1.30). De-
spite these differences, all the compsognathids fall
within the usual range of scapula/humerus ratio seen
in most coelurosaurs, in which the humerus is often as
long as, or longer, than the scapula (Holtz et a/., 2004).
The most significant exceptions, remote from the con-
dition seen in Scipionyx, are in fact some advanced
coelurosaurs, such as the dromaeosaurids (0.63-Xu et
al., 1999; 0.83-Burnham et al., 2000; 0.68-Hwang e?
al., 2002).

Scapula - The scapula of Scipionyx is typical of a
theropod in being strap-like: the slender scapular blade,
6-7 times longer than it is deep, has parallel cranial and
caudal edges for most of the length up to the beginning
of the acromion. The blade of the right scapula opens up
into a forked dorsal end on account of the detachment of
a fragment from the cranial margin. Of the left scapula,
only the acromion emerges, between the coracoids and
the furcula. Therefore, the following description refers
to the right scapula.

Dorsally, the scapular blade terminates in a pointed,
irregular margin which originated after post mortem
damage; it cannot be known, therefore, if originally
there was any terminal expansion or what the precise
length of the blade was. However, based on the pre-
served margins, we can exclude that the possible dor-
sal expansion was as wide as that in Ngwebasaurus
(de Klerk ef a/., 2000) and Juravenator, in which the
expansion is more than twice the width of the neck of
the blade. The dorsal third of the blade of Scipionyx is
slightly recurved. Ventrally, the blade widens gradually
towards the glenoid cavity, rather than having a con-
striction (neck). This seems to be directed caudally, as in
almost all other theropods (compsognathids included),

and differently to deinonychosaurs and birds, in which it
faces more laterally than caudally (e.g., Rauhut, 2003).
The glenoid cavity is not well-exposed, on account of a
partial superimposition of the proximalmost portion of
the humerus. However, the scapula of Scipionyx had to
form the cranial and dorsocranial margins of the cavity,
giving a contribution greater than that of the coracoid.
On the cranial edge, towards the coracoid contact, there
is a pronounced acromion, developed on the same plane
as the blade (i.e., not laterally everted, as in advanced
Maniraptora). The right acromion is very pronounced,
but less abruptly enlarged than appears in the pictures
at a first glance (Fig. 87). In fact, a thin patch of soft
tissue (muscle and ?connective tissue) has overlapped
the craniolateral margin of the scapular blade, but not
enough to completely hide the outline of the bone un-
derneath (see hatched line in Fig. 88). Under a similar
circumstance (another patch of soft tissue: Fig. 89), the
acromion of the left scapula appears even more squared
in shape and more pronounced. The reconstructed out-
line of the acromion of Scipionyx indicates that it is
prominent but smoothly merged into the cranial scap-
ular blade, as in Compsognathus (Peyer, 2006) and in
other compsognathids.

Like in the cranial portion of the coracoid, the surface
of the acromion of Scipionyx has a radially rugose tex-
ture and reinforced margins. Unlike in Compsognathus
(Peyer, 2006), the strongly convex cranial outline of the
two elements, still visible despite the superimposition of
the soft tissues, indicates the presence of a scapulocora-
coidal notch (Fig. 89). The scapulocoracoidal suture, the
caudal two-thirds of which are still clear, is sigmoidal in
shape and crosses the glenoid cavity caudally. Matching
the rib cage’s cross-section (see Dorsal Ribs), the lat-
eral surface of the scapula of Scipionyx seems flat rather
than laterally everted towards the coracoid contact as in
Velociraptor and other deinonychosaurs (Norell & Ma-
kovicky, 2004).

Although generally similar, the scapulae of the
compsognathids differ from each other in some as-
pects. The scapular blade of Compsognathus (Peyer,
2006) is not as slender and straight as that of Scipio-
nyx, and is slightly more recurved. Juravenator dif-
fers from Scipionyx in having a marked, diagnostic
narrowest portion, located at the neck of the scapula,
close to the acromion (G6hlich & Chiappe, 2006). In
Sinosauropteryx (Currie & Chen, 2001), the scapular
blade appears more elongate than in Scipionyx (length
almost equal to ten times the width), with the acro-
mion less pronounced. In Sinocalliopteryx (Ji et al.,
2007a) and Huaxiagnathus (Hwang et al., 2004) the
elongation is similar to that seen in Sinosauropteryx
(i.e., greater than in Scipionyx), but the acromion is
more prominent than that of Scipionyx, and triangular.
The basal tyrannosauroids 7anycolagreus (Carpenter
et al., 2005a), Dilong (Xu et al., 2004) and Guanlong
(Xu et al., 2006) have an even more prominent tri-
angular acromion. The former taxon also presents a
scapulocoracoidal notch, an even more elongated blade
and a narrow scapular neck. The basal ornithomimo-
saur Sinornithomimus (Kobayashi & Lii, 2003) has a
more slender scapula (length almost eleven times the
width), bearing an acromion similar in shape but less
developed than that of Scipionyx.

e
=
e)
pd
<
<
2
[2a]
Pe
©)
=
175)
3
°
n
n
<
(,)
el
<
©)
Li
<
E
D
ci
©)

=5mm.

Sn
ica

Scale bar =

nyx samniticus. Scala metr

amniticus

CS

,

1)

Scipio

imbs of Scipior
i

i anteriori

girdle and forel

Pectoral

7
Fig. 87 - Cinto pettorale e art

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

ZA Il

87

Fig. 88 - Line drawing of the bones illustrated in Fig. 87. See Appendix 1 or cover flaps for abbreviations.
Fig. 88 - Disegno al tratto delle ossa illustrate in Fig. 87. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

88 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 89 - Close-up of the furcula of Scipionyx samniticus. Scale bar = 1 mm. See Appendix 1 or cover flaps for abbreviations.
Fig. 89 - Particolare della furcula di Scipionyx samniticus. Scala metrica = 1 mm. Vedi Appendice 1 o risvolti di copertina per le

abbreviazioni.

Coracoid - Both coracoids are well-exposed in ven-
tral view, one opposite the other as in a mirror, with the
right one slightly overlapping the left one. Based on the
length of the furcula, which articulated with the acromion
of each scapula, the two coracoids had to be almost ad-
Jacent and very close to the medial sagittal plane in the
undisturbed, articulated skeleton. The right coracoid is
better preserved than the left one, which is crossed by two
parallel branches of the large calcite vein that runs from
the right ceratobranchial I to the right manus, passing also
through the caudal cervical region. The coracoid is fan-
shaped, semicircular, higher than long, and narrower at
the caudal edge, as in Compsognathus (Ostrom, 1978;
Peyer, 2006), Huaxiagnathus (Hwang et al., 2004) and
Sinosauropteryx (Currie & Chen, 2001). A raised border
marks the margins of both coracoids.

Left and right coracoid foramina are clearly visible.
The coracoid foramen, suboval in outline, opens at mid-
length of the craniocaudal axis and at about 1/3 width
from the scapulocoracoid suture. Several scars that rep-
resent muscle attachment sites are visible on the surface
of both coracoids. The protuberances often interpreted as
biceps tubercles in other theropods (but see Carpenter ef
al., 2005, for a different interpretation) emerge caudally
to the coracoid foramen in both coracoids, in the form of
digitiform reliefs, with the caudal margin more prominent
and rounded (Fig. 90). The caudoventral process extends
further caudally to the level of the glenoid cavity, as is
the case in tetanurans (Rauhut, 2003). Both extension and
outline of this caudoventral process, as well as of the gle-
noid cavity, are more clearly visible in the left coracoid,
the caudal portion of the right one being partly covered
by the proximalmost portion of the right humerus. The
caudoventral process appears triangular as in Compso-
gnathus (Peyer, 2006), but less pointed and developed.

The caudoventral process of the latter is not as pointed
and developed as that of Sinosauropteryx, which, in turn,
is comparatively less pointed and developed than that of
the basal tyrannosauroids (Xu et al., 2004, 2006).

The coracoid contributes to the glenoid with its cranio-
ventral margin; this is slightly less than that of the scapu-
la, which forms the cranial and craniodorsal margins (see
above). The glenoid portion of the coracoid is non-evert-
ed, without any neck marking a definite separation from
the perimeter of the bone. Compared to all other elements

Fig. 90 - Close-up of the right coracoid of Scipionyx samniticus. Scale
bar = 1 mm. See Appendix 1 or cover flaps for abbreviations.
Fig. 90 - Particolare del coracoide destro di Scipionyx samniticus. Scala

metrica = 1 mm. Vedi Appendice 1 o risvolti di copertina per le abbre-
viazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 89

of the forelimb, the coracoid of Scipionyx is smaller than
that of Compsognathus and Sinosauropteryx; for exam-
ple, in Scipionyx it measures less than half the length of
the humerus, whereas in Compsognathus it is more than
half the length of the humerus.

Furcula - A delicate, boomerang-shaped bone can be
seen leaning against the right scapula and coracoid, at
the level of the scapulocoracoid suture. By its size and
shape, it can be identified without doubt as the furcula
seen in cranial view (Fig. 89). The right ramus is super-
imposed onto the lateral portion of the scapula, and lacks
most of its distal extremity, of which only a thin frag-
ment can be seen bent back in a cranial direction. The
left ramus is complete and directed towards the left ac-
romion; it is crossed cranially by the already mentioned
calcite vein.

In addition to the avian theropods, furculae have been
documented in nearly all but the most basal theropods
(such as Eoraptor and Herrerasaurus), including coelo-
physoids (Tykoski et al., 2002; Carrano et al., 2005;
Rinehart et a/., 2007), spinosaurids (Lipkin ef al., 2007),
allosauroids (Chure & Madsen, 1996; Coria & Currie,
2006; Sereno e? al., 2008), compsognathids (Hwang et
al., 2004; Peyer, 2006; Ji et al., 2007a), tyrannosaurids
(Makovicky & Currie, 1998; Larson & Rigby, 2005;
Lipkin et al., 2007), therizinosauroids (Xu et al., 1999a;
Zhang et al., 2001; Kirkland et al., 2005), oviraptoro-
saurs (Barsbold, 1983; Barsbold ef al., 1990; Ji et al.,
1998; Clark et al., 1999, 2001; Lii, 2002; Osmòlska et al.,
2004), troodontids (Xu & Norell, 2004) and dromaeosau-
rids (Norell ef a/., 1997; Xu er al., 1999b; Burnham et al.,
2000; Hwang et al., 2002; Makovicky et al., 2005). The
paired elements labelled “cl” by GGhlich ef al. (2006:
fig. 1) in the compsognathid Juravenator, in our opinion
represent a broken furcula. However, taking as reference
the data collected by Claessens (2004: Appendix 1), this
specimen of Scipionyx is one of the few theropod fossils
with a simultaneous preservation of gastralia and furcula:
this has permitted clear distinction between these often
questioned elements. If our identification of the cranial-
most gastralial element as a chevron-shaped gastralium
is correct, there are only three other specimens having
a furcula preserved together with that element: A//osau-
rus DINO 11541, A/bertosaurus RTMP 91.36.500 and
Daspletosaurus NMC 11315. Following Larson & Rigby
(2005), an unambiguous furcula is presumably preserved
among the gastralia of 7yrannosaurus FMNH PR2081,
but see Brochu (2003) and Nesbitt et a/. (2009) for a dif-
ferent interpretation.

In Scipionyx, the furcula can be easily distinguished
from gastralia by the following morphological character-
istics: the furcula is larger, having a diameter that is at
least twice that of the largest gastralium (i.e., the chevron-
shaped gastralium); the rami of the furcula taper more
abruptly than those of the chevron-shaped gastralium,
whereas the other gastralia have either an almost con-
stant diameter or taper very gradually; the rami of the fur-
cula exhibit a sigmoidal curvature, whereas gastralia are
straighter; the furcula has a distinet pointed projection,
called the hypocleideum, which is absent in gastralia; the
preserved (left) extremity ofthe furcula of Scipionyx bears
a flat epicleideal facet (the epicleideum) for articulation
with the scapular acromion, which in three-dimensionally

preserved furculae is spatulate and sulcated by scars; gas-
tralia do not have epicleideal facets, even if the lateral and
the medial gastralia contact via a flat surface.

As just mentioned above, in Scipionyx the furcula is
larger and stouter than the cranial chevron-shaped gastra-
lium. According to Claessens (2004), in larger theropods
such as allosaurids and tyrannosaurids, the furcula is only
about 1/3 the size of a cranial chevron-shaped gastralium.

Besides the characters described above, the furcula of
Scipionyx has also the following features: based on Nes-
bitt et al. (2009), the general shape of the furcula is more
U-shaped than V-shaped, because, towards the symphysis,
the dorsal margins of its two rami are slightly concave,
rather than straight (Fig. 89); the epicleideal facets are
slightly flexed laterally, seem to be not much expanded
and taper to a point; the two rami, at least in proximity to
the symphysis, seem compressed craniocaudally in cross-
section, a character usually found in the Maniraptori-
formes; the hypocleideum of Scipionyx projects ventrally
and seems to belong entirely to the right ramus because
of the presence alongside it of a suture with the counter-
lateral element. The apex of the hypocleideum is covered
by a thin edge of ?connective tissue. Contra Nesbitt et al.
(2009), who estimated an interclavicular angle (i.e., the
angle between the rami of the furcula) of 140° for Sci-
pionyx, we obtain an angle not greater than 125° with the
same method of measurement. In Huaxiagnathus (Hwang
et al., 2004), the compsognathid fossil in which the fur-
cula is most clearly preserved, the bone resembles that of
Scipionyx in being U-shaped and in having an interclavic-
ular angle of 130°, but differs in that it lacks a distinct
hypocleideum. In coelophysoids, the furcula is variably
U- or V-shaped and has an angle of 115-140°; Suchomi-
mus (Lipkin et al., 2007) has a 111° V-shaped furcula; the
furcula is V-shaped also in allosauroids and ranges from
120° to 135° (Nesbitt er al, 2009); in tyrannosaurids,
the furcula is U-shaped with laterally flexed epicleideal
facets and a clavicular angle ranging from 71 to 113°
(Nesbitt e? al., 2009); in therizinosaurids, the U-shaped
furcula ranges from 135° to 160° (Nesbitt ef al., 2009);
in the Maniraptora, the furcula is generally U-shaped and
bears an angle of approximately 80-90°, but can reach
60° in most advanced non-Ornithurae forms (Nesbitt et
al., 2009) and 103° in Velociraptor, which also presents
a short hypocleideum (Norell & Makovicky, 2004). An
incipient hypocleideum is preserved in troodontids (Xu
& Norell, 2004).

Forelimb

Both forelimbs are articulated and almost complete
in Scipionyx. The elements of the right arm are visible
in their entirety, with the proximal and median segments
exposed, respectively, in laterocaudal and in lateral view;
the manus is visible in dorsomedial view. Of the left side,
only the manus is fully exposed, in palmar view (Figs.
87-88). Scipionyx displays also the very rare and superb
fossilisation of the horny talons present on the tip of each
digit. Only the distal portion of the second right ungual
phalanx is lost, as a result of a crack in the supporting slab
of Plattenkalk that presumably formed during collection.
The phalangeal count is 2-3-4. The unguals have a rela-
tively low curvature and bear sharp horny talons.

90 CRISTIANO DAL SASSO & SIMONE MAGANUCO

As stated by Dal Sasso & Signore (1998a), Scipionyx
is characterised by elongate raptorial forelimbs, measur-
ing about 48% of the presacral length (skull included).
They are comparatively longer than in most other non-
avian coelurosaurs except dromaeosaurids, which have
by far the longest forelimbs among non-avian theropods
(humerus and manus longer than in Scipionyx). The rela-
tive length of the forelimb is mostly due to elongation of
the manus. In Scipionyx, the manus measures 42% of the
whole forelimb length; therefore, it approaches the range
(43-50%) found in other compsognathids (Currie & Chen,
2001; Hwang et al., 2004; Peyer, 2006; Ji et al., 2007a;
2007b) and matches the ratio of several deinonychosaurs
and basal birds in which the manus constitutes about 40-
45% of the forelimb length, irrespective of age (e.g., Xu
et al., 1999; Burnham et al., 2000; Elzanowski, 2002;
Russell & Dong, 1993; Norell & Makovicky, 2004; Xu &
Norell, 2004). The manus is 34% of the whole forelimb
length in the basal ornithomimosaur Sinornithomimus
(Kobayashi & Lii, 2003), 30% in the basal oviraptoro-
saur Caudipteryx (Zhou et al., 2000), whereas, just like in
Scipionyx, it 1s 42% in the basal alvarezsaur Ngwebasau-
rus (de Klerk et al., 2000).

The epipodials of Scipionyx are about half the length of
the long manus. In other words, the manus is about 80% of
the combined length of humerus and forearm. According
to Padian (2004), a value between 66% and 75% is typical
of Coelurosauria, but there are many examples of higher
ratios, especially among composgnathid coelurosaurs. In
Sinosauropteryx (Currie & Chen, 2001; Ji ef al/., 2007b),
the manus is 84-91% of the combined length of humerus
and radius. In Sinocalliopteryx (Ji et al., 2007a) and Hua-
xiagnathus (Hwang et al., 2004), the manus is as long
as the humerus plus the radius, whereas in German and
French Compsognathus the ratios are 74% and 85%, re-
spectively (Peyer, 2006). Similar values are measured
also in basal tyrannosauroids (e.g., 91%-Carpenter et al.,
2005a; 80% based on the ulna, which includes the olecra-
non-Xu et al., 2006), in Ngqwebasaurus (73%-de Klerk
et al., 2000) and in basal troodontids (e.g., 83%-Russell
& Dong, 1993; Xu & Norell, 2004). Underestimated but
still comparable values, based on the ulna (including the
olecranon process) are available for dromaeosaurids (e.g.,
62%-Xu et al., 1999; 66%-Burnham et al., 2000) and basal
birds (64%-Elzanowski, 2002). A shorter manus is, instead,
present in the basal ornithomimosaur Sinornithomimus
(51%-Kobayashi & Lii, 2003) and the basal oviraptorosaur
Caudipteryx (43%-Zhou et al., 2000).

Humerus - The right humerus is exposed in caudola-
teral view (Figs. 87-88). In this aspect, the epiphyses and
their articular surfaces are clearly visible. The outline of
the proximal epiphysis is an evenly convex curvature
which delineates the humeral head and continues caudally
in the internal tuberosity. This tuberosity is placed dis-
tally about 13% of the way down the length of the shaft.
The cartilaginous area capping the humeral head is clearly
visible under UV light (see Appendicular Articular Carti-
lages). The margin of the head appears well-marked by a
raised, robust border.

The deltopectoral crest of the right humerus is directed
craniomedially, dipping into the matrix, so it is difficult to
evaluate its morphology. However, following the outline
of the left humerus, exposed in medial view, the delto-

pectoral crest of Scipionyx appears moderately developed.
It extends up to about one-third of the length of the hu-
merus, and expands the width of the proximal end of the
humerus up to twice its minimum shaft diameter. In our
opinion, the deltopectoral crest of the humerus of Scipio-
nyx resembles that of the other compsognathids, and the
differences mentioned by Peyer (2006) are mostly due
to different degrees of exposures of the humeri. A pos-
sible exception is Sinosauropteryx, which has a relatively
short and massive humerus, with a powerful deltopectoral
crest, extending for more than half the length of the bone
(Currie & Chen, 2001; Ji et a/., 2007b). In the tyrannosau-
roids, on the other hand, the proximal half of the humerus
is narrow, with a proximal end that is only slightly dif-
ferentiated (e.g., Xu et al., 2004; Carpenter ef al., 2005a).
It is slender and gracile, with a weak deltopectoral crest
and a strong spherical head, in the basal ornithomimosaur
Sinornithomimus (Kobayashi & Lii, 2003). It is robust,
with a strong deltopectoral crest, in the basal alvarezsaur
Nqwebasaurus (de Klerk et al., 2000).

Under grazing light, an apparent bump crossing the
diaphysis of the right humerus of Scipionyx, and a couple
of other bumps under the coracoids as well, which cause
alteration of the relief of the bones, can be seen. Those
bumps mark, respectively, the position of the diaphysis
and the proximal epiphysis of the left humerus (Figs. 88,
91), a position that is confirmed by CT scan imaging (Fig.
92). The bump corresponding to the diaphysis of the left
humerus tapers distally towards the epiphysis. The, left
distal epiphysis, which emerges between the 4° and 5"
dorsal ribs, is not expanded and shows only the radial
condyle. This indicates that the left humerus lies in medial
view. On the other hand, the distal epiphysis of the right
humerus gradually expands from the diaphysis to form

Fig. 91 - Oblique view under grazing light of the humeri of Scipionyx
samniticus, showing the elongate bump caused by the diaphysis of the
left humerus undercrossing the right one. See Appendix 1 or cover flaps
for abbreviations.

Fig. 91 - Vista obliqua e in luce radente degli omeri di Scipionyx sam-
niticus, che mostra la gobba allungata causata dalla diafisi dell’omero
sinistro nel punto in cui passa sotto il destro. Vedi Appendice 1 o risvolti
di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 9 |

è

i,
di BEI
A 17000008,

Fig. 92 - Computed tomography image of Scipionyx samniticus, showing the left side of the specimen, confirms the position of the left
humerus (red line) with respect to the right one (green line), as inferred from observations under grazing light.

Fig. 92 - Tomografia computerizzata di Scipionyx samniticus che mostra il lato sinistro dell’esemplare e conferma la posizione
dell’omero sinistro (linea rossa) rispetto al destro (linea verde), come si deduce dalle osservazioni in luce radente.

the distal condyles. Of these, the ulnar condyle appears
slightly more developed (Figs. 87-88). The entepicondyle
lies on the craniomedial side that faces towards the slab
and, so, is not visible.

The humerus of Scipionyx, especially in its distal half,
is rather straight if compared to the typically tetanurine
sigmoid shape. As expected, its curvature resembles
much more that of the other compsoganthids (Currie &
Chen, 2001; Hwang ef al., 2004; Gòhlich & Chiappe,
2006; Peyer, 2006) and, more in general, of the other non-
maniraptoran coelurosaurs (e.g., Kirkland ef al., 1998;
Kobayashi & Li, 2003; Carpenter et a/., 2005a, 2005b)
except Coelurus (Carpenter ef al., 2005).

Radius - The right radius is clearly exposed in dorsal
view (Figs. 88, 93); only a small portion of the proximal
epiphysis is not visible. It is a straight, slender bone, cy-
lindrical in shape and with a constant transverse diameter,
as is the case in other compsognathids (e.g., Gòhlich &
Chiappe, 2006), Coelurus (Carpenter et al., 2005b) and
Ornitholestes (Carpenter ef al., 2005b). Although the
shape of the radius in the Compsognathidae is not diag-
nostic, the radius of Scipionyx clearly appears different
from that of some basal tyrannosauroids, which is flat-
tened and expanded distally (Xu et a/., 2004, 2006), or
bowed (Carpenter e? a/., 2005a). The bony wall of the dia-
physis is partially collapsed, revealing its thinness and the
presence of a hollow interior.

The left radius is exposed in ventral view; its distal
epiphysis and two short tracts of the diaphysis are vis-
ible, the shortest in the spatium interosseum of the right
forearm, and the other one close to the right elbow. Here,
the diaphysis is split into two longitudinal sections, whose
displacement clearly reveals the mentioned hollow inte-
rior. A small triangular splinter of the proximal epiphysis
of the left radius emerges between the diaphyses of the
two humeri. Based on the combined observation of the

distal epiphyses of the right and left radii, the articular
surface for the carpals appears more expanded dorsally
than ventrally (Figs. 88, 93).

The radius of Scipionyx is about 60% the length of the
humerus. It is 57-61% in Sinosauropteryx (Currie & Chen,
2001), 62% in Huaxiagnathus (Hwang et al., 2004), 64%
in the German Compsognathus (Ostrom, 1978), 71% in

Fig. 93 - Radii, ulnae and carpi of Scipionyx samniticus. Scale bar
=2 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 93 - Radii, ulne e carpi di Scipionyx samniticus. Scala metrica
=2 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

92 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Juravenator (G6hlich & Chiappe, 2006) and 79% in the
French Compsognathus. Similar ratios are found in many
other coelurosaurs, such as the basal tyrannosauroids
Guanlong (68%-Xu et al., 2006), Coelurus (68%-Carpen-
ter et al., 2005b) and Tanycolagreus (72%-Carpenter et
al., 2005a), the basal ornithomimosaur Sinornithomimus
(75%-Kobayashi & Li, 2003), Ornitholestes (68%-Car-
penter et al., 2005b), Nqwebasaurus (76%-de Klerk et al.,
2000), Microraptor (78%-Hwang et al., 2002) and Sinor-
nithoides (71%-Russell & Dong, 1993). More elongate
radii are found in other Maniraptora, such as Mei (93%-
Xu & Norell, 2004).

Ulna - Like the radius, the ulna of Scipionyx is a slen-
der, cylindrical element (Figs. 88, 93). Its proximal epi-
physis is more robust, whereas the distal epiphySsis, as far
as can be seen in the available views, appears not at all
expanded. The right ulna, through a faint concave medial
arch, produces an apparent spatium interosseum. Parts of
the proximal articular surface and of the olecranon are
hidden under the right humerus and the cranialmost intes-
tinal loop. As in the right radius, the diaphysis is fractured
and hollow.

The diaphysis, the whole distal epiphysis and the
proximal portion of the proximal epiphysis of the left
ulna, lying in ventral view, are partly visible. In fact, the
left ulna, after passing under the right ulna, radius and
humerus, emerges in the thoracic region, caudally to the
end of the 4° dorsal rib, still in articulation with the hu-
meral condyle. Here, it shows a moderately pronounced
olecranon process.

If we follow Peyer (2006) in considering the differ-
ence in height between the radius and the ulna as the
height of the olecranon, the olecranon of Scipionyx is as
well-developed as in the German Compsognathus (Os-
trom, 1978). In Sinosauropteryx (Currie & Chen, 2001),
it is more developed. On the other hand, Huaxiagnathus
(Hwang et al., 2004), Sinocalliopteryx (Ji et al., 2007a)
and Juravenator (GGhlich & Chiappe, 2006) have a small,
relatively short olecranon process, as is the case also in
the basal tyrannosauroids (Carpenter e? a/., 2005a; Xu et
al., 2006). In the basal ornithomimosaur Sinornithomimus
(Kobayashi & Lii, 2003), the olecranon is even more re-
duced and the distal expansion is delicate, to00.

The distal epiphysis of the ulna is abruptly and great-
ly expanded in the French Compsognathus, but such a
marked difference with Scipionyx is partly due to differ-
ent exposure (cranial view in the former taxon). By the
way, according to Peyer (2006) Compsognathus differs
from Scipionyx (Dal Sasso & Signore, 1998a) in having
a straighter, not at all bowed, ulnar shaft. In our opinion,
this difference is not so relevant and may well be a con-
sequence of the different views in which the bones are
exposed. As a matter of fact, in Scipionyx itself the left
and right ulnae appear straight and slightly bowed, re-
spectively.

The ulna of Sinosauropteryx (Currie & Chen, 2001)
is even more robust and shorter, with a clearly expanded
distal epiphysis. In contrast, in Coe/urus (Carpenter et al.,
2005b) the ulna is more slender than in Scipionyx, with
a very thin shaft and a distal epiphysis that is gradually
expanded in all the views.

In Scipionyx, as in Huaxiagnathus (Hwang et al.,
2004), radius and ulna are subequal in transverse diameter

for almost all their length. Unlike Scipionyx, the trans-
verse shaft diameter of the ulna is slightly greater than
that of the radius in Compsognathus (Peyer, 2006); the
difference is even more pronounced in Sinosauropteryx
(Currie & Chen, 2001), although not to the degree of ad-
vanced coelurosaurs such as Microraptor (Hwang et al.,
2002), in which the bowed ulnar shaft is approximately
twice the thickness of the radius.

Finally, with regard to the length of the ulna in relation
to the humerus, in Scipionyx the former is about 70% the
length of the latter. The value is 62% in Huaxiagnathus
(Hwang et al., 2004), about 77% in the German Comp-
sognathus (Ostrom, 1978) and Sinosauropteryx (Currie
& Chen, 2001), and 90% in the French Compsognathus.
Among non-compsognathid coelurosaurs, the value is 76-
77% in the basal tyrannosauroids (Carpenter ef a/., 2005a;
Xu et al., 2006), about 75% in the basal ornithomimosaur
Sinornithomimus (Kobayashi & Li, 2003), 76% in the
basal alvarezasaur Ngwebasaurus (de Klerk et al., 2000),
78% in Sinornithoides (Russell & Dong, 1993), 82-90%
in dromaeosaurids (Xu et a/., 1999; Burnham et al., 2000;
Hwang et al., 2002) and about 90% in Archaeopteryx (El-
zanowski, 2002).

Carpus - As noted by Dal Sasso & Signore (1998a),
the carpus of Scipionyx is composed of two elements only
(Figs. 93-95): a small, simple lenticular-shaped bone is
more clearly exposed in the right carpus, preserved in
contact with the radius on both arms — we interpret this
proximal medial carpal bone as the radiale; another bone
with a more complex shape and contacts is located distally
to the radiale — and interpreted herein as a distal carpal.

In the right wrist, where it is exposed in dorsal view,
the distal carpal is a flattened and sigmoid element, deep-
est proximodistally in its medial half; in the left wrist, it
appears in ventral view and is deeper in its lateral half.
This pattern reflects the intimate articular contact with the
neighbouring bones, which leaves no space for additional
cartilaginous elements: proximally, the distal carpal con-
tacts the radiale for half of its articular surface, then it
reaches the radius and with its lateralmost extremity the
ulna; distally, it caps the proximal end of metacarpal I
plus the medial half of the proximal end of metacarpal
II — the proximal end of mcll is moderately convex with
a slight central elevation in ventral view, which appears
to subdivide its articular surface in two halves. Looking
at the differences in thickness and shape of the two dis-
tal carpals preserved in two opposite views in the right
and left wrists of Scipionyx, it is apparent that the volume
of this bone is complementary to that of the radiale and
the proximal articular facet of the first metacarpal (mel).
We interpret this carpal element as the first distal carpal
(del), which, as in other theropods, overlaps the proximal
articular facet of metacarpal I and part of the proximal
articular facet of metacarpal II.

There is no other carpal bone in the two wrists of Sci-
pionyx. According to Gauthier (1986), in absence of any
other distal element, a single distal carpal indicates the
fusion of del and dc2 during the ontogeny. However, a su-
ture is not detectable on the surface of neither the right nor
the left distal carpal 1 of Scipionyx, even at high magnifi-
cation (Fig. 95). It is interesting to note that Chure (2001)
verified that in A//osaurus the morphology of the carpals
is consistent throughout the ontogenetic series, and that

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 93

Fig. 94 - Metacarpals and manual phalanges of Scipionyx samniticus. Scale bar = 5 mm. See Appendix 1 or cover flaps for
abbreviations.
Fig. 94 - Metacarpali e falangi delle mani di Scipionyx samniticus. Scala metrica = 5 mm. Vedi Appendice 1 o risvolti di coper-
tina per le abbreviazioni.

the line of fusion between del and dc2 is still visible as
a groove in all but the two largest specimens examined
by him. Therefore, based on the fact that Scipionyx is a
hatchling (see Ontogenetic Assessment), the fusion be-
tween del and de2 would have occurred very precocious-
ly during ontogeny, probably already during embryogeny.
According to this hypothesis, the adult morphology of the
carpus would have been acquired very early and main-
tained unchanged up to adulthood, without any further
substantial modification, although it must be taken into
consideration that studies on several specimens of A//o-
saurus (Holtz et al., 2004) and Falcarius (Zanno, 2006)
demonstrated a certain individual/ontogenetic variability
in the degree of carpal fusion that renders the assessment
of a carpal formula, and the homology of carpal elements,
complicated. The other possible explanation for the ab-
sence of any suture is that the distal carpal of Scipionyx
derives from a single centre of ossification. Without going
further into the discussion of the homology of this ele-
ment, we remind the reader that autoradiographic studies
of the development of the carpus in the chicken indicate a
single centre of ossification for the avian semilunate car-
pal (Chure, 2001).

The radiale in Sinosauropteryx is more flattened and
comparatively smaller than that in Scipionyx. In the for-
mer, it is positioned below the middle of the distal end
of the radius, rather than being shifted apart towards the
medial side of the wrist. A further difference is in the dis-
tal contact: according to Currie & Chen (2001), in Sino-
sauropteryx it articulates not only with a large distal car-

pal, but also with the first metacarpal. In Compsognathus
(Peyer, 2006) and Huaxiagnathus (Hwang et al., 2004),
the radiale spans almost the entire width of the distal ra-
dius, being more similar to that of Scipionyx in its me-
diolateral extension. In the tyrannosauroid Guan/ong, the
radiale is instead considerably larger, being mediolater-
ally more extended than the distal carpal (Xu et a/., 2006
— suppl. info: fig. 2f-g).

The “semilunate carpal” of the Maniraptora (e.g.,
Rauhut, 2003; Chure, 2001; Elzanowski, 2002) has a
bulky trochlea and is markedly crescent-shaped. In Scipio-
nyx, in contrast, in both the available views the distal car-
pal appears somewhat flattened, without any bulky proxi-
mal trochlear surface, and, thus, results definitely different
from the true semilunate carpals (sensu Chure, 2001). In
our opinion, the similar flattened morphology of the dis-
tal portion of the carpus visible in the French specimen
of Compsognathus may well represent its original, natural
morphology, without the need for supposing that it was
originally semilunate and underwent considerable flatten-
ing during diagenesis, as hypothesised by Peyer (2006).

In the French Compsognathus, according to Peyer
(2006), the elements distal to the radiale are two sepa-
rated, but tightly bound together, elements that might
be fused to form the dcel+dc2 block. Peyer (2006) also
observed, in the German Compsognathus, three isolated
elements comparable in shape to the three carpals found
in the French specimen, contra Ostrom (1978), who sug-
gested that there were probably only two elements in the
wrist of the German Compsognathus. A total of five carpal

94 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 95 - Close-ups of the carpal bones of Scipionyx samniticus. On the surface of the left (A) and right (B) distal carpal element, no
suture is detectable. Scale bar = 0.5 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 95 - Particolari delle ossa carpali di Scipionyx samniticus. Sulla superficie dell’elemento carpale distale sinistro (A) e destro (B)
non è individuabile alcuna sutura. Scala metrica = 0,5 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

elements has been reported for Sinosauropteryx (Currie &
Chen, 2001). However, we express some doubt about the
association to the same (right) wrist of the two elements
tentatively identified by Currie & Chen (2001) as addi-
tional carpals, because of their underlying position and
uncertain identification: they might simply be the coun-
terlateral left carpals, disarticulated phalanges or other
indeterminate bones. Whatever the number of carpals, in
Sinosauropteryx the distal carpal is broad and low, just
like in Scipionyx, and extends beyond the first metacarpal
to contact metacarpal II; unlike Scipionyx, it caps only the
lateral half of the first metacarpal, the medial one being
occupied by a peculiar proximomedial flange of mcl. A
small, disk-like element, identified as the ulnare by Cur-
rie & Chen (2001: fig. 8a), can be seen between the ulna
and the second metacarpal of Sinosauropteryx. In Sci-
pionyx, the lateral portion of the distal carpal reaches the
same position, so we suggest that the large distal carpal
and the ulnare described by Currie & Chen (2001) might,
in fact, be the still unfused dcl and dc2. This might be the
case also for the holotype of Huaxiagnathus (Hwang et
al., 2004), where a purported “semilunate carpal” — which
we think is much too small to be this bone in actual fact
— contacts laterally a very small “ulnare”, positioned be-
tween radius and ulna and close to the middle of the ar-
ticular surface of mell.

Among compsognathids, two stacked, plate-like car-
pals are described only in Huaxiagnathus and Scipionyx.
The exposure of the left and right wrists in the holotype
of Huaxiagnathus (Hwang et al., 2004: fig. 8) is the same
as that of Scipionyx and allows direct comparisons. No-
tably, the shape and size of the radiale and the dc1+dc2
(“semilunate carpal” in Hwang et a/., 2004) are similar
in both genera: in dorsal view, the radiale is thin and me-
diolaterally expanded, whereas in ventral (palmar) view,
it is shorter and bulkier proximodistally; the distal carpal
has a complementary shape in both views. In Huaxiagna-
thus, the presence of an additional small, ossified carpal
between the ulna and melll, identified as a dc3 by Hwang
et al. (2004), would be the only noticeable difference with
Scipionyx. A total of four elements is, in fact, reported in
the former as well as in Sinocalliopteryx (Ji et al., 2007a):
given their comparable position, they are probably ho-
mologous to each other.

The tyrannosauroid 7anycolagreus (Carpenter et
al., 2005a) exhibits a carpal formula identical to that of
Scipionyx, with a proximal carpal and, functionally, a
single distal carpal derived from the fusion of the distal
carpals. The proximal carpal, described as “semilunate”
but termed “radiale” by Carpenter ef al. (2005a), is a
little more expanded proximodistally and considerably
larger than the radiale of Scipionyx, its transverse diam-
eter being greater than that of the distal portion of the
radius (Carpenter ef al., 2005a: fig. 2.11 I). The distal
carpal is as flattened and elongate as the one of Sci-
pionyx, but it is larger, overlapping mel, mell and, al-
though only barely, mceIll too. Tanycolagreus and Hua-
xiagnathus are, therefore, the theropod taxa most similar
to Scipionyx with regard to carpal morphology. A cer-
tain resemblance can be seen also with the distal carpal
of Coelurus, described by Rauhut (2003) as “somewhat
rectangular or flattened rather than truly semilunate in
ventral (palmar) view”.

In the adult specimen of the tyrannosauroid Guanlong
(Xu et al., 2006), the distal carpal is much deeper than
in Scipionyx, possessing a transverse trochlea proximal-
ly and a semilunate shape in ventral view. Its position
(contacting mel and half of mell) is similar to the condi-
tion seen in A/losaurus, Scipionyx and oviraptorosaurs,
whereas in therizinosauroids, troodontids, dromaeosau-
rids and basal birds it articulates primarily with mell. The
basal ornithomimosaurs are considerably different from
Scipionyx: no stacked elements are present, and the car-
pals are represented by three or more disk-like elements,
one being a well-developed ulnare, all aligned on the
same plane (Kobayashi & Lii, 2003). In Ngwebasaurus
(de Klerk ef a/., 2000), two carpals are preserved, ten-
tatively identified as dcl and dc2. Sinornithoides (Rus-
sell & Dong, 1993) is described in having a carpus com-
posed of two elements: a radiale and a semilunate. This
count is reminiscent of Scipionyx, but in the former the
radiale contacts laterally also the ulna and, in all likeli-
hood, the semilunate is larger and crescent-shaped as in
other deinonychosaurs and, more in general, in most of
the Maniraptora. As a matter of fact, in the closely re-
lated Mei the manus is fossilised folded in avian fashion
(Xu & Norell, 2004), a position probably not feasible for
Scipionyx.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 95

Metacarpus - The metacarpus of Scipionyx is com-
posed of three elements and, on the whole, it appears
moderately slender (length/width ratio of mel-III = 2) and
not as narrow and elongate as in Ornitholestes and most
of the Maniraptora (e.g., Rauhut, 2003). The metacarpus
is compact, with mell and mellI closely appressed in lat-
eral contact (Figs. 88, 94). On the other hand, contrary to
Sinosauropteryx (Currie & Chen, 2001), the lateral mar-
gin of mel in Scipionyx, well-visible in the right element,
is concave. As a consequence, the contact between mcl
and mcell would have been partly limited, although not as
limited as in Deinonychus (Gishlick, pers. comm., 2000).
The right metacarpus, exposed in a view that is mainly
dorsal but permits to see the medial side, shows a few
marked extensor pits. The left metacarpus is exposed in
medioventral (mediopalmar) view, allowing observation
of the condyles.

In Scipionyx, mel is 40% of the length of mell and
twice as long proximodistally than wide transversely,
resulting narrower than the bulky, squared mcl of Sino-
sauropteryx (Currie & Chen, 2001). McI is 40-50%
of the length of mcell also in Sinosauropteryx (Cur-
rie & Chen, 2001), Juravenator (GG6hlich & Chiappe,
2006), Sinocalliopteryx (Ji et al., 2007a), Huaxiagna-
thus (Hwang et al., 2004), basal tyrannosauroids (Xu
et al., 2004: fig. 2; Carpenter et al., 2005a; Xu et al.,
2006, suppl. info: fig. 2f) and basal ornithomimosaurs
(Makovicky et al., 2004), whereas it is as long as mell,
or slightly shorter, in derived forms. It is a bit shorter
in Compsognathus (Gishlick & Gauthier, 2007; Pey-
er, 2006) and tends to become even shorter in some
Maniraptora (Norell & Makovicky, 1999; Russell &
Dong, 1993; Osmélska et al., 2004; Padian, 2004). On
the other hand, mcl is more elongate (63% of the length
of mcell) than that in Scipionyx in the basal alvarezsaur
Nqwebasaurus (de Klerk et al., 2000), in which mell is
notably more gracile at its mid-shaft than mel (58%). In
Scipionyx, the proximal end of mel, which is regular in
outline and lacks any trace of the proximomedial flange
visible in Sinosauropteryx (Currie & Chen, 2001), is not
perfectly aligned with the ones of mcell and melll; in
fact, it is slightly more distally positioned, as is the case
in many Tetanurae (Gishlick, pers. comm., 2000). On
the right mcl, a faint extensor pit can be seen; the distal
condyles are asymmetric, with the medial wider and lo-
cated more proximally than the lateral one. This asym-
metry matches the asymmetry of the proximal facet of
the phalanx I-1 and, in life, made the pollex divergent
from the other digits (see below), albeit not as much as
in Tugulusaurus (Rauhut & Xu, 2005), in which the me-
dial side is entirely situated more proximally than the
lateral one, and in advanced Maniraptora, in which the
asymmetry is more marked. Rauhut (2003) and Langer
(2004) reported that this condition is widespread in sau-
rischian dinosaurs.

Mcell is the longest bone in the manus of Scipionyx,
and its shaft is subequal in width to that of mel. Very
similar length and proportions can be seen in Huaxia-
gnathus (Hwang et al., 2004) and Juravenator (G6hlich
& Chiappe, 2006). Its cylindrical structure, without any
marked medial expansion in the proximal epiphysis,
is characteristic of the majority of coelurosaurs (e.g.,
Rauhut, 2003: Holtz et al., 2004). As mentioned in the de-
scription of the carpus, the proximal margin has a central

elevation in ventral (palmar) view: this indicates that the
articular surface is subdivided in two equally expanded
facets, one for the contact with the ulna and the other for
the contact with the distal carpal (Fig. 95A). On the dis-
tal epiphysis of the right mcll is an extensor pit that is
slightly more developed than the faint ones visible on mel
and mcelllI. Also, the medial collateral ligament fossa is
well-exposed (Fig. 96). According to Rauhut (2003), the
extensor pits are shallow or absent in the metacarpals of
most coelurosaurs.

In Scipionyx, melll is slightly thicker than half
mell, and it is slightly shorter than the latter, as in the
other compsognathids (Ostrom, 1978; Currie & Chen,
2001; Hwang e? al., 2004; Gòhlich & Chiappe, 2006;
Peyer, 2006). Contrary to mel and mell, which have ex-
panded ends (especially the distal ones), mcellI straight,
slender and unexpanded distally. In the Italian comp-
sognathid, however, it is not as slender as in Compso-
gnathus (Peyer, 2006; Gishlick & Gauthier, 2007), in
which it resembles the derived maniraptorans in its de-
gree of slenderness and in having the bone bowed later-
ally rather than being straight. In Sinocalliopteryx (Ji et
al., 2007a), as well as in the basal tyrannosauroid 7any-
colagreus (Carpenter ef al., 2005a), melll is even more
slender and also considerably shorter than in Scipionyx.

Fig. 96 - Close-up of the distal epiphysis of the 2°° right metacarpal
(dorsal view) of Scipionyx samniticus. Scale bar = 0.5 mm. See Appen-
dix 1 or cover flaps for abbreviations.

Fig. 96 - Scipionyx samniticus, particolare dell’epifisi distale del 2°
metacarpale destro (norma dorsale). Scala metrica = 0,5 mm. Vedi
Appendice 1 o risvolti di copertina per le abbreviazioni.

96 CRISTIANO DAL SASSO & SIMONE MAGANUCO

The fact that the proximal epiphysis of melll lies on that of
mcell in the left manus — exposed in ventral (palmar) view
— and under the one of mell in the right manus — exposed
in dorsal view — is not random. Rather, it reflects a natural
anatomical condition that is diagnostic for the Tetanurae
(Holtz et al., 2004). Finally, meIV is definitely absent in
Scipionyx, contrary to Guanlong and Ornitholestes (Xu
et al., 2006), which are the only known coelurosaurs to
preserve a reduced metacarpal IV.

Manual phalanges - The hands of Scipionyx have three
digits each: digit I is the shortest, digit II is the longest and
digit III is the most gracile. The proximal phalanges are
exposed in the same view as their metacarpals, whereas
the distal phalanges and, in particular, all the ungual pha-
langes are rotated to expose the medial side, with the ex-
ception of the left ungual I, which lies in lateral view. The
phalanges I-1, II-1, III-2 and III-3 of the right manus are
traversed by the same calcite vein that crosses the neck
and the left scapulocoracoid. This secondary mineralisa-
tion (see Diagenetic Formations Possibly Related To Soft
Tissues) has fractured and separated the bones with a 1
mm-wide gap.

The non-ungual phalanx of the first digit has a con-
dylar asymmetry analogous to that of mel and a slightly
twisted shaft (Figs. 88, 94). For these reasons, the first
digit of Scipionyx diverged from the other two, like in
many other theropods; this increased the spread of the fin-
gers and permitted these dinosaurs to grasp larger objects
during flexion. On the contrary, the non-ungual phalanges
of digits II and III have straight, definitely non-twisted,
shafts and aligned condyles that are equally developed in
a distal direction. From the combined observation of the
non-ungual phalanges of both manus, it can be inferred
that each phalanx bears ginglymoid condyles; moreover,
the distal condylar portion has lateral and medial collat-
eral ligament fossae, which are well-delimited, deep and
opened proximally.

Remarkably, the shaft diameter of manual phalanx
I-1 is subequal to the shaft diameter of the radius (Figs.
88, 94). In Compsognathus (Peyer, 2006), Juravenator
(Gòhlich & Chiappe, 2006), Huaxiagnathus (Hwang
et al., 2004) and Sinocalliopteryx (Ji et al., 2007a: fig.
3b), the proximal shaft diameter of manual phalanx I-1
is even greater than the minimum shaft diameter of the
radius, and the authors above regard this character as
diagnostic of the Compsognathidae. In Sinosauropte-
ryx (Currie & Chen, 2001), this condition is even more
emphasised: the first phalanx of digit I and the ungual
phalanx that it supports are massive, each being as long
as the radius and thicker than the shafts and the distal
ends of either the radius or the ulna. In Sinosauropteryx,
Huaxiagnathus and Sinocalliopteryx (Ji et al., 2007a),
phalanx I-1 (except unguals) is the longest, whereas
in Scipionyx, Compsognathus (Peyer, 2006) and the
basal tyrannosauroid Tanycolagreus (Carpenter et al.,
2005a), the longest phalanx is II-2. Unlike Sinosau-
ropteryx (Currie & Chen, 2001), phalanx I-1 of Scipio-
nyx is shorter than mell, and the first digit is not larger
than the forearm bones; rather, it equals the epipodials
in length and thickness. In Scipionyx, all the penultimate
phalanges are elongate: phalanx III-3 is longer than the
sum of III-1 and III-2, which are both very short; phalanx
II-2 surpasses II-1; similarly, the length of I-1, which is

not preceded by other phalanges, is more than twice that
of mel. These proportions are similar to those found in
the vast majority of coelurosaurs (Rauhut, 2003).

Peyer (2006) described two possible phalanges, II-
1 and III-2, of the left manus of Compsognathus. These
phalanges, especially the latter, are very short, measur-
ing only 25% and 10% of the length of melII, respec-
tively; in other compsognathids, the respective ratios for
these phalanges are 32% and 41% in Scipionyx, 30% and
38% in Sinosauropteryx, and 40% and 50% in Jurave-
nator (GGShlich & Chiappe, 2006). The odd ratio of the
alleged phalanx III-2 of Compsognathus arouses doubts
on the completeness of that element. By the way, Gish-
lick & Gauthier (2007) suggested that the third digit was
reduced, perhaps even non-functional, based on the over-
all morphology and slenderness of metacarpal III and the
possibly preserved phalanges, highlighting the fact that
there is no evidence on either specimen of Compsogna-
thus of a claw-bearing ungual phalanx. In any case, fol-
lowing the reconstruction by Peyer (2006: fig. 9), digit III
would be shorter than digit I. A digit III shorter than digit I
is present also in Sinosauropteryx (Currie & Chen, 2001),
Huaxiagnathus (Hwang et al., 2004) and Sinocalliopteryx
(Ji et al., 2007a), whereas in Juravenator, according to
Gohlich & Chiappe (2006, suppl. info, table I), digits I
and III are subequal in length, with the latter digit slightly
shorter than (95% of its length) the former.

Based on these comparisons, Scipionyx differs from
all known compsognathids in having a digit III that is
considerably longer (123%) than digit I. Among non-
maniraptoran coelurosaurs, a digit III longer (110%) than
digit I is present in the adult individual IVPP V 14531
of the basal tyrannosauroid Guanlong (Xu et al., 2006:
fig. 2b). Based on the proportions of digits I and III, the
manus of Scipionyx can be compared — although not from
a morphological point of view — to that of some ovirap-
torosaurs such as Citipati, Conchoraptor and Oviraptor
(e.g., Clark et al., 2001; Osmélska et al., 2004). Manual
digit I is reported to be shorter than digit III also in the
troodontid Jinfengopteryx (Ji et al., 2005): we regard
this feature as unclear, because it is difficult to confirm
and quantify that statement based on the published fig-
ures, and there are some inconsistencies in the table of
measurements in Ji et a/. (2005). Digits I and III are sub-
equal in several Maniraptora (e.g., Ostrom, 1969; Xu ef
al., 1999: fig. 2; de Klerk et al., 2000: fig. 2; Xu et al.,
2002b; Russell & Dong, 1993), whereas digit III is longer
than digit I (115%) in Velociraptor (Norell & Makovicky,
1999). Finally, the digits of Archaeopteryx (Elzanowski,
2002) resemble those of Scipionyx in their relative pro-
portions, digit Il being the longest, digit I the shortest and
digit III intermediate in length. However, based on other
sources (e.g., Wellnhofer, 1985: fig. 1; Elzanowski, 2002:
fig. 2.1; Paul, 2002: fig. 4.3; Padian, 2004: fig. 11.1), the
difference in length between digits I and III is negligible
as they appear almost subequal. Outside the Tetanurae,
a digit III longer (121%) than digit I is reported also in
Dilophosaurus (Welles, 1984).

In Scipionyx, the curvature of the bony claws (ungual
phalanges) is emphasised by the horny sheaths, which
are rarely preserved in fossils. The dorsal margin of the
ungual phalanges is uniformly convex up to the articu-
lar margin, without the dorsal lip found in oviraptoro-
saurs, deinoychosaurs and birds (Currie & Russell, 1988).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 97

The base of the bony claws (i.e., their articular surface) is
quite tall, similar to what is seen in Juravenator (GGhlich
& Chiappe, pers. comm., 2006). However, the transition
from the proximal articular surface to the main body of the
ungual is not as markedly distinct in Scipionyx as in sev-
eral other theropods (e.g., via a small, shallow transverse
groove [Rauhut & Xu, 2005]), especially ventrally. There,
close to the articular margin, a quite developed flexor tu-
bercle, that is however shorter than half the height of the
articular surface, gradually rises up (Figs. 88, 94). Thus,
Scipionyx is similar to the other compsognathids except
Compsognathus in having flexor tubercles taller than 1/3
of the height of the ungual articular facets. Even larger
flexor tubercles, i.e., that are more than half the height of
the articular facets, are instead present in deinoychosaurs
and birds (Rauhut, 2003).

There is some variation in the morphology of the flexor
tubercles in Scipionyx. In ungual I, seen in lateral and me-
dial views, the ventral (palmar) margin runs straight from
the flexor tubercle to the articulation. In unguals 2 and 3,
this margin is concave and, consequently, the tubercles
appear more distinct from the main body of the bones.
A similar variation seems to be present in the holotype
of Huaxiagnathus (Hwang et al., 2004: fig. 7B), which,
on the other hand, has unguals that are more curved and
more gradually tapering. In contrast, the first ungual in
Nedcobertia is more robust, with a more prominent flexor
tubercle (Kirkland et a/., 1998).

In Scipionyx, the ungual phalanx of digit I, which is
the only ungual preserved in medial and in lateral views,
has slightly asymmetric vascular grooves (Fig. 88): on the
medial side (right phalanx), the sulcus bifurcates immedi-
ately from its apical portion towards a proximal direction,
delimiting a thin ventral lip; on the lateral side (left pha-
lanx), the sulcus bifurcates where it approaches the half
length of the phalanx, and delimits a deeper ventral lip.
Vascular grooves bifurcating in the distal half of the bone
are present on the medial sides of the ungueal phalanges
of the second and third digit.

Concerning the relative size, in Scipionyx manual un-
guals I and II are subequal in size, whereas ungual III is
slightly smaller. In the compsognathids Sinocalliopteryx
(Ji et al., 2007a) and Huaxiagnathus (Hwang et al., 2004),
as well as in the basal tyrannosauroids Di/ong (Xu et al.,
2004: fig. 21) and Guanlong (Xu et al., 2006, suppl. info:
fig. 2d) and in Sinornithosaurus (Xu et al., 1999), the trend
is the same, but ungual III is considerably smaller. There-
fore, contra Hwang et al. (2004), the large second ungual
is not a diagnostic character of Huaxiagnathus. By com-
bining information from the French and German speci-
mens, Peyer (2006) reconstructed the ungual phalanx II
of Compsognathus as the largest, and the ungual phalanx
III as the smallest, as is the case in Velociraptor (Norell
& Makovicky, 1999) and Archaeopteryx (Elzanowski,
2002). In Juravenator (Gòhlich & Chiappe, 2006), the
three ungual phalanges share an autapomorphic shape in
being high proximally and abruptly tapering at the mid-
point, and differ from each other only in size, ungual I be-
ing the largest, and ungual III being the smallest. Similar
but more emphasised proportions can be seen in the basal
tyrannosauroid 7anycolagreus (Carpenter ef al., 2005a).
In the majority of theropods, ungual I is considerably
larger than ungual II, which is almost equal to ungual III
(Senter, 2007). This is especially true for Torvosaurus and

the spinosauroids, where ungual I is 2/3 the length of the
radius, and also for Ngwebasaurus (de Klerk et al., 2000),
where ungual I measures 3/2 the length of ungual III. Re-
markably, a similar but even more emphasised condition
is present in the compsognathid Sinosauropteryx (Rauhut,
2003). Therefore, Sinosauropteryx markedly differs from
Scipionyx: in the latter, the ungual I is 2/5 the length of
the radius; in the former, ungual I slightly surpasses the
length of the radius.

Apart from their size, the unguals of Scipionyx re-
semble those of Compsognathus and Juravenator in their
moderate degree of curvature and in the proportions of the
flexor tubercles. The ungual phalanges are gently curved
also in Nqwebasaurus (de Klerk et al., 2000), but they are
definitely more elongate than in Scipionyx. In ornithomi-
mosaurs (Makovicky et al., 2004), the flexor tubercle is
even less pronounced and the curvature is even less em-
phasised, especially proximally. In contrast, the unguals
of many other maniraptoriforms are often more curved
than in Scipionyx and bear stronger flexor tubercles.

Pelvic girdle

The pelvic girdle of Scipionyx is incomplete. Part of
the dorsal blade of the left ilium, as well as the proximal
portion of the ischia, broke away and were irremediably
lost during collection of the specimen, when one of the
three main cracks of the slab occurred (Figs. 9-10). How-
ever, most of the pelvic girdle is preserved and the bones,
although unfused, partly maintained the orientation they
had in vivo. Scipionyx possesses an orthopubic pelvis with
a dolichoiliac ilium and an estimated length of the ischi-
um 3/4 the length of the pubis (Figs. 97-98).

In all compsognathids the shaft of the pubis is straight
in lateral view, but forms different angles with the iltum
and the long axis of the body. It is propubic (i.e., it points
cranioventrally) in Compsognathus (Ostrom, 1978; Pey-
er, 2006) and Mirischia (Martill et al., 2000; Naish et al.,
2004), whereas it is almost orthopubic (i.e., almost verti-
cal) in Sinosauropteryx (Currie & Chen, 2001), Sinocal-
liopteryx (Ji et al., 2007a), Huaxiagnathus (Hwang et al.,
2004) as well as in the basal tyrannosauroid Guan/ong (Xu
et al., 2006: fig. 2e). In Ornitholestes (Carpenter et al.,
2005b), the pelvis is almost orthopubic because of a com-
bination of structures: the pubic peduncle points slightly
cranially, but the pubic shaft is recurved. Although the
shape of the single girdle elements is different, the basal
therizinosauroid Fa/carius (Kirkland et al, 2005) and
most oviraptorosaurs (Osmélska ef al., 2004) have also
an orthopubic pelvis. On the other hand, the pelvis is of-
ten opisthopubic (i.e., with a retroverted, caudoventrally
pointing pubis) in deinonychosaurs (e.g., Xu et al., 1999;
Burnham e? al., 2000; Hwang et al., 2002).

Ilium - Except for a thin fragment of the mediodorsal
edge of the blade, the left ilium is completely obscured by
the sacral bones and by the right counterlateral element.
The latter is fairly complete, just partly eroded along the
cranial margin and lacking a dorsolateral portion of the
preacetabular ala (Figs. 97-98).

The ilium of Scipionyx appears craniocaudally short
in comparison with total body length or femur length, but
comparable proportions (ilium shorter than femur) can be

CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 97 - Pelvic girdle and hindlimbs of Scipionyx samniticus. Scale bar = 5 mm
Fig. 97 - Cinto pelvico e arti posteriori di Scipionyx samniticus. Scala metrica = 5 mm

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

99

I il sac

| pua

cm ume — —

_

pufo

cn

Fig. 98 - Line drawing of the bones illustrated in Fig. 97. See Appendix 1 or cover flaps for abbreviations.
Fig. 98 - Disegno al tratto delle ossa illustrate in Fig. 97. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

100 CRISTIANO DAL SASSO & SIMONE MAGANUCO

seen also in other compsognathids, including Sinosau-
ropteryx (Currie & Chen, 2001), Compsognathus (Peyer,
2006) and the large Sinocalliopteryx (Ji et al., 2007a:
fig.1), and in a number of coelurosaurs (e.g., Xu et al.,
1999; Burnham et al., 2000; Hwang et al., 2002; Xu et al.,
2004). The preacetabular portion is as long and high as
the postacetabular portion is, as is the case in many other
theropods, including most coelurosaurs (e.g., Rauhut,
2003) and all other compsognathids except Huaxiagna-
thus (Hwang et al., 2004), which, similar to the tyranno-
sauroids (Brochu, 2003; Xu et al., 2004; Xu et al., 2006),
has a slightly longer postacetabular ala.

The dorsal margin of the ilium of Scipionyx is uniform-
ly convex, as in Miîrischia (Naish et al., 2004), whereas it
is straight in Sinosauropteryx (Currie & Chen, 2001). The
preacetabular ala of Scipionyx is dorsoventrally expand-
ed, forming a hooked process directed cranioventrally.
According to Rauhut (2003), the ilium has a dorsoven-
tral expansion of the preacetabular ala that is more-or-
less hooked in many neotheropods, with the exception of
the forms more closely related to birds. This expansion
is indeed well-marked in ornithomimosaurs (Makovicky
et al., 2004) and in tyrannosauroids (Holtz, 2004) ex-
cept Dilong (Xu et al., 2004), whereas in Ornitholestes
(Carpenter et a/., 2005b) and in some compsognathids,
it is either only faintly developed or absent. In Compso-
gnathus, the cranial process of the ilium is not expanded
dorsoventrally, and its ventral edge, although not well-
preserved, is straight (Peyer, 2006). In Sinosauropteryx
(Currie & Chen, 2001) and Huaxiagnathus (Hwang et al.,
2004), the ventral margin of the preacetabular ala curves
gently cranioventrally and, like in Compsognathus, does
not have a distinct hooked process. In Mirischia, the dor-
soventral expansion is well-developed but not hooked,
terminating “in a square-ended process that is directed
cranio-ventrally” (Naish ef a/., 2004). Thus, the only
compsognathid showing a condition very similar to that
of Scipionyx is Sinocalliopteryx (Ji et al., 2007a).

In Scipionyx, the preacetabular ala bears along its
craniodorsal margin a small, but well-incised semicircular
notch that, if compared to the eroded, irregularly indented
margin of the neighbouring areas, does not seem an arte-
fact of preservation (Fig. 99). As a matter of fact, a notch
or concavity is found in exactly the same position in the
ilia of some other basal coelurosaurs, especially tyranno-
sauroids such as Dilong (Xu et al., 2004), Guanlong (Xu
et al., 2006) and Stokesosaurus (Rauhut, 2003), and the
derived forms. This craniodorsal concavity, together with
a preacetabular hooked ventral projection (see above), is
often considered a synapomorphy of the clade Tyranno-
sauroidea (Holtz, 2004). A notch is present also in Orni-
tholestes (Carpenter et al., 2005b: fig. 3.10A), where it
faces cranially rather than craniodorsally.

Unlike Sinosauropteryx (Currie & Chen, 2001), Sci-
pionyx has a supracetabular crest, albeit a weak one. Cra-
nially, it parallels the concave outline of the acetabulum,
whereas caudally it becomes straight and joins the ventral
margin of the postacetabular blade. The right ilium has
also two subparallel ridges, oriented vertically on the sur-
face of the iliac blade dorsal and caudal to the acetabulum.
Nevertheless, these ridges are not true iliac structures —
none of them, for example, is homologous to the “promi-
nent medial vertical crest” typical of the tyrannosauroids
(see Xu et a/., 2006) — but rather are due to the presence

Fig. 99 - Close-up of the craniodorsal margin of the preacetabular
ala of the right ilium of Scipionyx samniticus, where a small concav-
ity is found. Scale bar = 1 mm. See Appendix 1 or cover flaps for
abbreviations.

Fig. 99 - Scipionyx samniticus. Particolare del margine craniodor-
sale dell’ala preacetabolare dell’ileo destro, dove si trova una pic-
cola concavità. Scala metrica = 1 mm. Vedi Appendice 1 o risvolti
di copertina per le abbreviazioni.

of a sacral vertebral centrum underneath, which during
diagenetic crushing was sandwiched between the left and
right ilia (see Sacral Vertebrae).

The right ischial peduncle of Scipionyx is partly cov-
ered cranially by the right femur, whereas the left one is
indistinctly visible. As far as can be seen, the squared end
of the ischial peduncle extends ventrally as much as the
pubic peduncle, as seems to be the case in Compsognathus
(Peyer, 2006: fig. 7) and Ornitholestes (Carpenter et al.,
2005b). In Juravenator (G6hlich & Chiappe, 2006), the
ischial peduncle is more expanded ventrally, whereas in
Sinosauropteryx the pubic peduncle is only slightly lon-
ger than the ischial one (Currie & Chen, 2001), as is the
case also in the basal tyrannosauroid Guan/ong (Xu et al.,
2006) and in the basal ornithomimosaur Sinornithomimus
(Kobayashi & Lii., 2004).

In Scipionyx, Compsognathus, Huaxiagnathus
(Hwang et al., 2004) and Sinosauropteryx (Currie &
Chen, 2001), as well as in most tetanuran theropods
(Rauhut, 2003; Holtz ef a/., 2004), the pubic pedunele
is very large in lateral aspect, and craniocaudally ex-
panded, whereas the ischial peduncle is fairly small. In
contrast, Juravenator (Gòhlich & Chiappe, 2006) has a
well-developed ischial peduncle.

In Scipionyx, the cranial and the caudal margins of
the pubic peduncle are distinctly concave, and the ven-
tral margin has a low central peak, so the whole process
appears fan-like in lateral view. In Mirischia, Juravena-
tor (Gòhlich et al., 2006: pl. 7, fig. 2), the basal tyran-
nosauroids Stokesosaurus (Rauhut, 2003) and Guanlong
(Xu et al., 2006), and Ornitholestes the pubic peduncle
is expanded cranially, so that its cranial margin is con-

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 101

cave as well. Because of the mentioned low central peak,
the articular surface of the pubic peduncle in Scipionyx is
gently convex, like in Mîrischia (Naish et al., 2004).
Caudal to the ischial peduncle, the medial wall of the
brevis fossa (1.e., the “medial blade” sensu Currie & Zhao,
1993a) appears in lateral view as a robust triangular ala,
strengthened by ridged buttresses. The lateral wall of the
brevis fossa consists of a marked lateral ridge that caudal-
ly joins the ventral margin of the postacetabular ala. The
postacetabular ala terminates in a caudal truncation. The
ilium is caudally squared (i.e., “truncated”’) in many thero-
pods, with the exception of Ornitholestes and the Manirap-
tora (Rauhut, 2003), in which it tapers, and Huaxiagnathus
(Hwang et al., 2004), in which it is gently rounded.

Pubis - The right and left pubes run coupled perpen-
dicularly to the long axis of the ilium and parallel to the
right femur, very close to its cranial margin (Figs. 97-98).
Except for the ischial process, hidden by the right femur,
the right pubis is well-exposed in lateral view. The fact
that the dorsal extremity of its apron is exposed, partial-
ly covered by the intestine, suggests that this thin bony
lamina bent caudally during sliding of the right pubis
onto, and cranially towards, the left one. In contrast, the
apron of the left pubis bent cranially, and now emerges
obliquely from the plane of the fossiliferous slab as a tall,
thin crest running along nearly half the length of the entire
pubis. A photograph taken with a grazing view (Fig. 100)
reveals that the top of the crest divides the cranial face
and the caudal face of the apron itself. Part of the me-
dial side of the left pubis emerges to the left of the caudal

Fig. 100 - Grazing view of the pubic bones of Scipionyx samniticus.
The top of the crest, indicated by the white line, divides the cranial face
from the caudal face of the left pubic apron. See Appendix 1 or cover
flaps for abbreviations.

Fig. 100 - Vista radente delle ossa pubiche di Scipionyx samniticus. La
sommità della cresta indicata dal connettore divide la faccia craniale
del grembiule pubico sinistro da quella caudale. Vedi Appendice 1 o
risvolti di copertina per le abbreviazioni.

face: it is marked by a deep groove, by a further change of
plane and by a more rugose texture, and overlaps the shaft
of the right femur for a short tract. The left apron does
not extend ventrally up to the distal end of the pubis. The
ventral interruption of the left apron indicates the pres-
ence of an elongate, oval opening (pubic foramen) that
was located ventral to the conjoined aprons of the pubes
and dorsal to their distal ends. Such an opening is present
in the Ceratosauria and the vast majority of the Tetanurae
(Rauhut, 2003), including compsognathids (e.g., Naish et
al., 2004) but not Ornitholestes (Carpenter et al., 2005b).
According to Martill ef a/. (2000), this opening may have
accommodated a ventral pneumatic duct leading to a post-
pubic air sac.

The pubis of Scipionyx measures about 2/3 ofthe length
of the femur, like in the basal troodontid Sinornithoides
(Russell & Dong, 1993). The length of the pubis is ap-
proximately that of the femur in Compsognathus (Ostrom,
1978; Peyer, 2006), Coelurus, Ornitholestes (Carpenter
et al., 2005b) and Sinornithomimus (Kobayashi & Li,
2004); is 3/4 and 4/5 the length of the femur respectively
in the specimens NIGP 127586 (Currie & Chen, 2001: fig.
la) and NIGP 127587 (Currie & Chen, 2001: fig.1b) of
Sinosauropteryx; and is 3/4 the length in Sinocalliopteryx
(fig. 3c) and Dil/ong (Xu et al., 2004: suppl. info).

Proximally, the pubis contacts the ilium through an ex-
panded, well-rimmed iliac process, which shows a faintly
concave articular margin complementary to the fan-like
pubic peduncle of the ilium. Proximocaudally, the pubis
of Scipionyx tapers to form an ischial process that, al-
though covered by the right femur, can be seen with a CT
scan (Fig. 101). The ischial process seems quite short and
rectangular in outline, and incised ventrally by an obtura-
tor notch that is very similar in size to that of Compso-
gnathus (Peyer, 2006).

As mentioned above, the pelvis of Scipionyx is ortho-
pubic: the pubes form a right angle with the long axis of
the body. The pubic shafît is definitely slender, straight in
lateral view and, based on the observable relief, is also
rod-shaped in cross-section. It terminates in a distinct
foot with a finely pitted surface. The right pubic foot,
previously thought to be hidden by the right femur (Dal
Sasso & Signore, 1998a: fig. 2), is, in fact, entirely ex-
posed but difficult to see except under proper light orien-
tation (Fig. 102) because it is tightly appressed onto the
lateral wall of the femur. As first suggested by Auditore
(pers. comm., 2008) during preparation of his drawings
for this monograph, the low bump seen at the same level
under the right femur represents the underlying left pubic
foot (Figs. 98, 102), a fact confirmed by CT scan. The
pubic foot has the shape of a wooden golf-club and lacks
the cranial process; caudally it projects to form a pro-
cess four times longer craniocaudally than the minimum
craniocaudal diameter of the pubic shaft. In the compsog-
nathids Compsognathus (Peyer, 2006), Huaxiagnathus
(Hwang et al., 2004), Sinocalliopteryx (Ji et al., 2007a),
Mirischia (Naish et al., 2004) and Aristosuchus (Naish et
al., 2001), the pubic foot also projects only caudally, but
it is definitely longer (i.e., half or more than half as long
as the pubic shaft) and more pointed than in Scipionyx,
in which the limited development of this bone is prob-
ably due to its earlier ontogenetic stage (see Ontogenetic
Assessment). Sinosauropteryx represents an exception
among compsognathids (Currie & Chen, 2001): in this

102 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 101 - Computed tomography of the pelvic region of Scipionyx samniticus. A parasagittal slice between the two femora, virtually
cutting the right ischium (red) and the right pubis (blue), reveals that the exposed portion of the obturator process continues into an

axe-shaped lamina. See Appendix 1 or cover flaps for abbreviations.

Fig. 101 - Tomografia computerizzata della regione pelvica di Scipionyx samniticus. Una fetta parasagittale tra i due femori, che taglia
virtualmente l’ischio destro (rosso) e il pube destro (blu), rivela che la porzione esposta del processo otturatore continua in una lamina
a forma di scure. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

taxon, the pubic foot expands caudally from the shaft of
the pubis but bears also a moderate cranial expansion, as
seen in basal tyrannosauroids Di/ong (Xu et al., 2004),
Tanycolagreus (Carpenter et al., 2005a) and Coelurus
(Carpenter e? a/., 2005b), in which the pubic foot is in any
case definitely larger (more than half the pubic length).
The cranial expansion is larger in Guanlong (Xu et al.,
2006), advanced tyrannosauroids and ornithomimosaurs
(e.g., Kobayashi & Lii, 2004). In Ngwebasaurus, the pu-
bis expands equally cranially and caudally into a small,
mediolaterally narrow pubic foot with a very indistinct
outline (de Klerk e? a/., 2000), whereas in oviraptorosaurs

and therizinosauroids it has a well-developed cranial pro-
cess and a caudal process of variable length (Osmòlska e?
al., 2004). In deinonychosaurs and birds, the caudal pro-
Jection of the pubic foot is again reduced to a round knob
or is completely absent (e.g., Xu et a/., 1999; Burnham
et al., 2000; Elzanowski, 2002; Russell & Dong, 1993;
Makovicky & Norell, 2004).

The minimum diameter of the pubic shaft equals the
minimum diameter of the ischium, as is the case in Sinor-
nithomimus (Kobayashi & Lii, 2004). The shaft of the
ischium is more slender than the pubic shaft in Huaxia-
gnathus (Hwang et al., 2004: fig. 9), Compsognathus

Fig. 102 - Photographs taken in perpendicular (A) and grazing (B) view of the right pubic foot and knee joints of Scipionyx samniticus.
Scale bar = 2 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 102 - Scipionyx samniticus. A) piede pubico destro; B) vista radente dello stesso e delle articolazioni delle ginocchia. Scala metrica
=2 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY l 03

(Peyer, 2006: fig. 7), Sinosauropteryx (Currie & Chen,
2001: fig. 10) and, probably, Sinocalliopteryx (Ji et al.,
2007a: fig. 3c), and is even more slender in Ornitholestes
(Carpenter ef a/., 2005b), where it is about half the diam-
eter of the pubic shaft. On the other hand, the ischial shaft
is more robust than the pubic shaft in the basal tyranno-
sauroid Guanlong (Xu et al., 2006), and is even more ro-
bust in deinonychosaurs and birds (e.g., Russell & Dong,
1993; Norell & Makovicky, 2004; Padian, 2004), where
the diameter of the ischial shaft is up to twice the diameter
of the pubic shaft.

Ischium - The distal halves of both ischia are exposed,
parallel to one another and projecting caudoventrally at an
angle of about 54° with the pubic shafts (Figs. 97-98). The
shaft of the right ischium, interrupted by a vertical fracture
that occurred in the fossiliferous slab, continues cranially
to the fracture overlapping onto the medial surface of the
left femur. Here, as in the mid-distal portion of the ischial
shaft, the surface ofthe ischium can be easily distinguished
from the surface of the femur because of the smoother tex-
ture of the preserved periosteum (see Periosteal Remains).
In this area, the ischium abruptly widens to form an obtu-
rator process that served as an attachment for the ischial
head of the M. puboischiofemoralis externus (Hutchinson,
2001). In Scipionyx, the distal portion of the obturator pro-
cess, which is not covered by the right femur and is more
clearly visibile under UV light (Fig. 103), appears squared
and forms a right angle with the shaft of the ischium rather

Fig. 103 - The pelvis of Scipionyx samniticus under ultraviolet-induced
fluorescence. See Appendix 1 or cover flaps for the abbreviation.

Fig. 103 - Il cinto pelvico di Scipionyx samniticus visto in fluorescenza
indotta da luce ultravioletta. Vedi Appendice 1 o risvolti di copertina
per l’abbreviazione.

than a ventrodistal notch. The ventrodistal notch is absent
also in Sinocalliopteryx (Ji et al., 2007a), Huaxiagnathus
(Hwang e? al., 2004), Compsognathus (Peyer, 2006), Di-
long (Xu et al., 2004: fig. 2), Ornitholestes (Carpenter et
al., 2005b: fig. 3.10A), the oviraptorosaurs (Osmolska et
al., 2004), the therizinosauroids (Kirkland ef al, 2005)
and the dromaeosaurids (Xu et a/., 1999; Burnham ef al.,
2000; Hwang e? al., 2002). Unlike Scipionyx, however, in
all these taxa the obturator process gradually rises from the
shaft of the ischium and forms a pointed triangular process.
This triangular morphology, resulting from the absence of
the notch at the distal base of the obturator process, has
been considered by Rauhut (2003) to be a reversal of a
derived character within theropods. The scenario is com-
plicated by Sinosauropteryx, in which a triangular mor-
phology and the notch are present (Currie & Chen, 2001),
and by Mirischia (Naish et al., 2004), in which a triangular
morphology and the notch are present in the right ischium
but not in the left one.

The obturator process of Scipionyx, as can be seen
from CT scans, is not triangular, even proximally: apart
from the squared distal base, its proximal outline is defi-
nitely axe-shaped (Fig. 101). An obturator notch opens
proximally, separating the cranial margin of the obturator
process and the ischial process. A similar obturator notch
is found in Compsognathus and in most coelurosaurs. Un-
like Scipionyx and other tyrannosauroids (and more gen-
erally the Coelurosauria), but like more basal tetanurans,
in Guanlong (Xu et al., 2006) the notch is replaced by
a relatively large foramen. The compsognathid Mirischia
asymmetrica, as its name celebrates, has a pelvic foramen
on the left ischium and a notch on the right element.

Distally, the ischia of Scipionyx are expanded into two
small, cranially hooked ischial feet, the craniocaudal di-
ameter of which is twice the minimum craniocaudal di-
ameter of the ischial shafts. The ischial foot of the Italian
compsognathid is very similar to that of Compsognathus
(Peyer, 2006), Sinosauropteryx NIGP 127587 (Currie &
Chen, 2001: fig. 10), Huaxiagnathus (Hwang et al., 2004),
Sinocalliopteryx (Ji et al., 2007a), Ornitholestes (Carpen-
ter et al., 2005b) and, possibly, Dilong IVPP V 11579 (Xu
et al., 2004: fig. 2j), albeit slightly more expanded in some
of these taxa (e.g., Compsognathus). The ischial foot is
also similar but variably expanded in ornithomimosaurs
(Rauhut, 2003), whereas, unlike Scipionyx, the ischium
tapers distally to a point in the compsognathid Mirischia
(Naish et al., 2004) and in many Maniraptora and tyran-
nosaurids (Hutchinson, 2001; Rauhut, 2003).

The ischial feet of Scipionyx have a finely pitted tex-
ture and raised borders. A flat facet (Fig. 104) is exposed
on the medial side of the left one; in life, this facet prob-
ably formed a symphysis with the opposite (right) facet,
but it is not possible to assess if they remained unfused
during ontogeny or later co-ossified, possibly in relation
to gender. Some soft tissue remains are preserved on the
caudal tip of the left ischial foot of Scipionyx, as well
as cranially and caudally to both shafts (see Pelvic And
Hindlimb Muscles).

Except for the well-exposed distal end, the left ischi-
um is less complete than the right one. Its caudal margin,
marked distally by a small fracture, can be followed up
to the large crack in the slab, where it becomes vertical.
The medial wall of the bone is lost, unveiling its internal
aspect, just right to this vertical caudal margin. A possible

104 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 104 - Close-up of the ischial feet of Scipionyx samniticus. Scale bar
=1 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 104 - Particolare dei piedi ischiatici di Scipionyx samniticus. Scala
metrica = 1 mm. Vedi Appendice 1 o risvolti di copertina per le abbre-
viazioni.

proximal portion of the left ischium can be seen between
the 5° sacral centrum and the medial wall of the brevis
fossa (Fig. 105). It consists of a knob-like caudal bump,
possibly representing the contact with the ilium, and a free
cranial margin, that we tentatively regard as the proximo-
medial contribution to the acetabulum. We consider this
proximal portion as being the left ischium, also because
it is in ideal continuity with the vertical caudal margin of
this bone and lies on a similar sagittal plane. In fact, it is
covered either by the centrum of SS or by the intestine.
Measured from the iliac suture to the distal end, the
ischium of Scipionyx is estimated to be about 75% the

Fig. 105 - Close-up of a bone here interpreted as the only exposed
proximal portion of the left ischium of Scipionyx samniticus. Scale bar
=1 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 105 - Scipionyx samniticus. Particolare di un osso qui interpre-
tato come l’unica parte esposta della porzione prossimale dell’ischio
sinistro. Scala metrica = 1 mm. Vedi Appendice 1 o risvolti di copertina
per le abbreviazioni.

length of the pubis, as is the case in Ornitholestes (Car-
penter ef al., 2005b), Falcarius (Kirkland et al., 2005:
fig. 1g) and Sinornithomimus (Kobayashi & Lii, 2004).
The ischium is slightly longer in Guanlong (80%-Xu et
al., 2006: fig. 2e), slightly shorter in Sinocalliopteryx
(70%-Ji et al., 2007a), Sinosauropteryx (66-70%-Cur-
rie & Chen, 2001), Huaxiagnathus (66%-Hwang et al.,
2004), Compsognathus (64%-Peyer, 2006), most orni-
thomimosaurs (around 60%-Makovicky et al., 2004) and
Caudipteryx (60%-Zhou et al., 2000: tab. 1), and much
shorter in deinonychosaurs (more or less about 50% the
length of the pubis [Norell & Makovicky, 1997; Xu et al.,
1999; Burnham ef al., 2000; Hwang et al., 2002; Russell
& Dong, 1993]).

Hindlimb

The crack crossing the pelvic girdle of Scipionyx con-
tinues all along the caudal margin of the left femur and, in
the end, transversely cuts the bones of the crus (tibiae and
fibulae) not distant from the knee joint. In all likelihood,
that crack is responsible for the irremediable loss of the
distal portions of the hindlimb when the calcareous slab
was collected. Thus, the length of the tibia and fibula can
be estimated only (see Skeletal Reconstruction And...).
Whatever the length of the crus, the hindlimbs would have
been well-developed, as indicated by the femur/humerus
ratio (143%). Like in the bones of the forelimbs, the thin
walls of the femora are partly collapsed in some areas,
revealing spacious, hollow internal cavities. In Scipionyx,
the craniocaudal length ofthe proximal end ofthe fibula is
about 85% the craniocaudal length of the proximal end of
the tibia and, thus, falls within the range of values (equal
or greater to 75%) for Coelurosauria (Holtz et a/., 2004).

Femur - The thigh bone is the largest preserved ele-
ment of the skeleton of Scipionyx (Figs. 97-98). It is al-
most straight, being slightly bowed cranially, like in other
compsognathids (Currie & Chen, 2001; Hwang et al.,
2004; Naish et al., 2004; Peyer, 2006; Ji et al., 2007a)
and Ornitholestes (Carpenter et al., 2005b), but unlike
Coelurus (Carpenter et al., 2005b) and the dromaeosau-
rids (e.g., Burnham er al., 2000), in which the shafît is
strongly bowed cranially and medially. In the right femur,
the greater trochanter clearly emerges just below the su-
pracetabular margin of the ilium, but the lateral aspect of
the bone allows us to follow only its dorsal outline. The
lateral aspect precludes also any information on the femo-
ral head. The lesser (=cranial) trochanter is placed proxi-
mally at a lower level, but it is well-developed dorsoven-
trally and craniocaudally into a squared end and seems to
be separated from the femoral shaft by a narrow cleft. The
lesser trochanter can be defined as “alariform” (‘“wing-
like” in Gauthier, 1986) because, in dorsal view, its end
has a lenticular shape, is mediolaterally flattened and is
craniocaudally elongate. A similarly developed, plate-like
lesser trochanter, separated from the greater trochanter
by a narrow cleft, is present in Compsognathus (Peyer,
2006), Juravenator (G6hlich & Chiappe, 2006), Huaxia-
gnathus (Hwang et al., 2004), Sinosauropteryx (Currie
& Chen, 2001) and in a number of basal tetanurans and
basal coelurosaurs (Rauhut, 2003). It markedly differs
from the condition seen in most Maniraptora, in which

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 105

it is fused to the very well-developed greater trochanter,
forming a high crest. No accessory trochanter, homolo-
gous to that ofthe ornithomimosaurs, present more or less
in the vicinity of the cranial base of the cranial (lesser)
trochanter (e.g., Kobayashi & Barsbold, 2005), is visible
in Scipionyx. In contrast, a trace of this accessory tro-
chanter is visible in the compsognathids Mirischia (Naish
et al., 2004) and Aristosuchus (Naish et al., 2001), which
exhibit a peculiar bifid, craniodorsally directed trochan-
ter: this trochanter was regarded by Hutchinson (2001b)
as a fused accessory and lesser trochanter, and accord-
ing to him it probably represents a primitive feature for
coelurosaurs and not a shared derived similarity between
these two compsognathids. An accessory trochanter distal
to the lesser trochanter is reported also in the tyranno-
sauroid Guanlong and in some oviraptorosaurs and dro-
maeosaurids (Xu et a/., 2006). In Scipionyx, the caudola-
teral trochanter seems to be absent, as in Compsognathus
(Peyer, 2006). Because of their caudomedial orientation,
the fourth trochanters, if well-developed, should be vis-
ible in the right and left femora. No trace of this fourth
trochanter can be seen on the right femur of Scipionyx,
whereas the left femur is covered by the ischial shaft
in that area. Therefore, we cannot establish whether the
fourth trochanter was feebly developed or totally absent.
The fourth trochanter is absent in Compsognathus (Peyer,
2006), coded as absent or greatly reduced in Juravenator
(Gohlich & Chiappe, 2006), present as a low ridge (i.e.,
greatly reduced) in Coelurus (Carpenter et al., 2005b),
Mirischia (Naish et al., 2004) and Aristosuchus (Naish
et al., 2001), and strongly reduced to a weakly developed
crest or a slight depression, or even entirely absent, in
Ornitholestes (Carpenter et al., 2005b), Tanycolagreus
(Carpenter ef al., 2005a) and most of the Maniraptori-
formes (Rauhut, 2003). The fourth trochanter is reported

Fig. 106 - Close-up of the crura of Scipionyx samniticus. The arrow
points to the caudal cleft separating the fibular condyle from the inter-
nal condyle and the main body of the tibia. Scale bar = 2 mm. See
Appendix 1 or cover flaps for abbreviations.

Fig. 106 - Particolare delle crura (gambe) di Scipionyx samniticus. La
freccia indica la fenditura caudale che separa il condilo fibulare dal con-
dilo interno e dal corpo principale della tibia. Scala metrica = 2 mm.
Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

to be present, but was not described, in Sinosauropteryx
(Currie & Chen, 2001), and is prominent in Di/ong and
Guanlong (Xu et al., 2006).

Distally, the femoral epicondyles of Scipionyx slightly
protrude in respect to the caudal margin of the diaphy-
sis. Comparing the left medial epicondyle with the left
and right lateral epicondyles, the latter appear larger than
the former, the ratio of the craniocaudal extension of the
lateral to medial one being 4/3. The marked separation
between medial and lateral epicondyles, well-visible in
the distal epiphysis of the left femur (which is exposed
in mediocaudal view), suggests that the popliteal fossa is
open distally (Figs. 106-107).

Tibia - Of the tibiae of Scipionyx, only the proximal
epiphyses and probably less than half of the diaphyses are
preserved, exposed mostly in lateral (right) and medial
(left) views (Figs. 97-98). The right tibia is also partly
rotated, showing a small portion of its caudal wall. The
cnemial crest of Scipionyx is weakly developed, as in
most small-bodied tetanuran theropods, including comp-
sognathids (Currie & Chen, 2001; Hwang et al., 2004;
Géohlich & Chiappe, 2006; Peyer, 2006; Ji ef al., 2007a).
The fibular condyle is well-distinct and well-developed:
cranially, as is the case in tetanurans, including many coe-
lurosaurs (Rauhut, 2003), it is separated from the cnemial
crest by a narrow and deep notch, the incisura tibialis
(Figs. 102B, 106); caudally, as in most theropods (Rauhut,
2003), it is separated from the internal condyle and the
main body of the tibia by a deep caudal cleft (Figs. 106-
107). The latter is visible in the form of a well-excavated
sulcus also in the left tibia.

Fig. 107 - Close-up of the proximal epiphyses of the left tibia and fibula
of Scipionyx samniticus. The arrow points to the caudal cleft shown
also in Fig. 106. Scale bar = 2 mm. See Appendix 1 or cover flaps for
abbreviations.

Fig. 107 - Particolare delle epifisi prossimali di tibia e fibula sinistre di
Scipionyx samniticus. La freccia indica la fenditura caudale mostrata
anche in Fig. 106. Scala metrica = 2 mm. Vedi Appendice 1 o risvolti di
copertina per le abbreviazioni.

106 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fibula - Like the tibiae, the fibulae of Scipionyx pre-
serve only their proximal epiphyses and a short tract of
their diaphyses (Figs. 97-98). This is particularly true for
the left fibula, which is preserved for a distance less than
half that of the right one. The fibula is slender and medio-
laterally compressed, the lateral surface is convex, and the
dorsal margin is equally convex and seems to be able to
accommodate an extensive contact with the fibular con-
dyle of the tibia. Among compsognathids, a comparable
concavity is present in Compsognathus (Peyer, 2006),
whereas it seems particularly shallow in Sinosauropteryx

(Currie & Chen, 2001). However, possible diagenetic de-
formation of the curvature of the fibular condyle leads us
to consider this condition in Scipionyx with caution. On the
left fibula, a fossa can be seen on the medial surface near its
proximal end (Fig. 107): therefore, as in many theropods
(Rauhut, 2003), except, for instance, Coelurus (Carpenter
et al., 2005b), the concave dorsal margin seen on the right
element would continue in this medial fossa. The marked,
gradual distal tapering of the preserved right fibula sug-
gests that the missing diaphyseal portion of this bone was
very slender, as is often the case in coelurosaurs.

Table 1 - Selected numbers and measurements (in mm) of the skeleton of the holotype of Scipionyx samniticus.
Symbols: — approximate measurement; - measurement not possible; ? incomplete or partially covered element. Where not specified,
height or width or diameter are taken perpendicular to the length. Diameter refers to both craniocaudal and mediolateral measurements,
or to any intermediate plane, according to the exposed view of each element.

General Right Left

Preserved body length

(premaxilla to Ca9) OSTO

Estimated body length
(whole tail included)

461.0

Presacral length (skull included)

164.7

Presacral length

113.0

Cervical length 44.0

Dorsal length 69.0

Estimated sacral length 23.0

Preserved caudal length 50.5

Gleno acetabular length
(at acetabular cranial margin)

Forelimb length
(humerus to end of phalanx II-3)

Manual digit I length 10% 14.8

Manual digit II length 218.6 250

Manual digit III length 21751 18.3

Skull Right | Left

Skull length
(premaxilla to quadrate)

Maximum skull length SI

Skull height (frontal to jugal) 28.5

Reconstructed skull height above
orbit

Antorbital length
(premaxilla to lacrimal)

Supratemporal fenestra length

Supratemporal fenestra height 4.0 -
Infratemporal fenestra length 6.3 -
Infratemporal fenestra height 12.8 -
Orbit length Lila] -
Orbit height 18.2 -
Antorbital fenestra length di, | -,
Antorbital fenestra height 8.6 -
Maxillary fenestra length 17 -
Maxillary fenestra height 4.3 -
Promaxillary fenestra length 0.7 -
Promaxillary fenestra height 1.5 -
Apertura nasi ossea length 6.3 -
Apertura nasi ossea height |

Frontal length

Frontal maximum width

Nasal length

Prefrontal length

+

Lacrimal length of horizontal ramus

Lacrimal height of vertical ramus

Maxilla length

Maxilla height

}

Pterygoid length

Number of scleral plates

Mandible Right Lett

Hemimandible length - 47.3

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 107

Estimated hemimandible length 444 C6 A SI 6.7
at glenoid fi si

CH 4.9 6.1
Reconstructed hemimandible -19
maximum height (above surangular) o C8 54 6.0
Dentary length 29.3 - C9 -5.0 6.0
Dentary height at mid-length 412 i C10 4.9 7.0
Ceratobranchial I length ISURMe2180 Dorsal vertebrae

length= exposed length at centrum

Length | Height

Ceratobranchial I minimum 0.7 -0.8 height= height of neural arch
mid-shaft diameter i i (top of spine to base of arch)

DI 5.1 6.4
Teeth Right Left D2 4.9 | 254
Upper tooth row length 18.4 - IDE È 6.0
Lower tooth row length 19.5 - D4 | 24.6 | 5.3
Premaxillary tooth row length 4.4 - DS ngn) | DIR
Maxillary tooth row length 13.0 123 D6 26.3 6.9
Number of upper teeth (pm+m) 12 (5+7) D7 500 25.0
Number of lower teeth 10 D$8 -5.9 | 24.5
Largest pm tooth (pm2) 30 i D9 6.0 7.4
crown height i

D10 6.5 -6.8
Largest pm tooth (pm2) 1.3 a
fore-aft basal length j DII - 74.9
Largest m tooth (m4) crown height 4.0 73.8 DIZ 6.3 5.8
Largest m tooth (m4) 1.5 n D13 6.3 6.7

fore-aft basal length
Sacral vertebrae

Largest dentary tooth (d3) 23 length= exposed length at centrum
crown height i height= height of neural arch
(top of spine to base of arch)

Largest dentary tooth (d3) 13 a I

fore-aft basal length SI 7.4 8.1
Serrations per mm in maxillary 11-12 S2 - -
teeth (distal carina) 3 ; :
E i; mm in dentary teeth 13 i; ; Pra

SS SA 24.8

Axial skeleton Caudal vertebrae

Cervia vench e spose ngi scena | Len Hei
Pope a I Length | Height (top of spine to base of arch) |

(top of spine to base of arch) Cal | 5.4 4.9 |
CI - 5.6 Ca2 5.4 5.4
69) 4.8 7.4 Ca3 5.6 | 4.7
3 4.3 Sa Ca4 SÒ 4.2
‘cd 4.5 5.8 Cas 7.4 | 5
CS 4.7 6.2 Ca6 6.6 4.8

CRISTIANO DAL SASSO & SIMONE MAGANUCO

108
Ca7 6.8 4.5
Ca8 TIRA 4.1
Ca9 - Sh)
Length | diameter
Chevron 4 8.1 0.7
Chevron 5 6.7 0.7
Cervical ribs linear length Right Left
Cri SD Ol
(ro 14.3 159
Cr4 292 23.0
Co 26.0 -
Cr6 76.0 -
Cri Isso Al22
Cr8 92 TO
Cr9 12.8 (241
Crl0 78.9 21.6
Dorsal ribs linear length Right Left
Drl 15.6 Ha
Dr2 DIR 210.7
Dr3 PISA 25.4
Dr4 SPAAÌ 2.235
Drs 221.4 217.8
Dr6 30.6 19%
Dr7 30.0 220101
Dr8 24.3 -
Dr9 NS 20
Drl0 215.9 2129
Drll 215.4 211.4
Dr12 153 dels:
| Dr13 20.7 -
Sacral ribs length Right Left
Srl È :
| Sr2 sz È
\Sr3 © E

Furcula width

Furcula angle between two rami

Humerus length

Sr4 46 PSE
Sr5 -6.4 -
Gastralia | Right | Left
Chevron-shaped gastralium 28.50
linear length i
Chevron-shaped gastralium 0.48
maximum shaft diameter ì
Cranial medial gastralia linear 9.50 Ù
length (best exposed element) i
Cranial medial gastralia
maximum shaft diameter VU -
(best exposed element)
Cranial lateral gastralia 10.30 i
linear length (same row) i
Mid-medial gastralia 920 o)
linear length (best exposed element)

î
Mid-medial gastralia
maximum shaft diameter Ve) -
(best exposed element)
Mid-lateral gastralia 9
linear length (same row) ‘03, L
Caudal medial gastralia 7.80 È
linear length (best exposed element) i
Caudal medial gastralia
maximum shaft diameter 0.26 -
(best exposed element)
Caudal lateral gastralia
linear length (same row) Ù ‘
Pectoral girdle and forelimbs Right Left |
Scapula length 23.8 -
Scapula maximum dorsal width 4.3 -
Scapula minimum midshaft width 3.0 -
Scapula maximum ventral width -8.4 -
Coracoid width 10.8 11.4
Coracoid height 6.8 6.8

Humerus proximal diameter

Humerus minimum
mid-shaft diameter

Humerus distal diameter

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

| Radius length

2o:l VEE

Radius minimum 19 ì
| mid-shaft diameter di
o: length 717.8 19.3
Ulna minimum mid-shaft diameter 1.8 -
Proximal carpal diameter bei 71.4
Distal carpal diameter 3.0 24
© NA width (mcel-III) DR 4.9
pure! I length 4.0 4.0
Metacarpal J minimum 17 91,5
| mid-shaft diameter

Metacarpal II length 10.6 10.6
i ianete Ai, EC
Metacarpal III length SH 8.6
pizzi mis | nta
Manual phalanx I-1 25 25
proximal diameter y
Wisuual phalanx I-1 length 9.5 “Al
Manual phalanx I-2 (ungual) N) 26.6
maximum length i
Manual phalanx II-1 length -6.7 15
Manual phalanx II-2 length 10.2 10.4
Manual phalanx II-3 (ungual) 91.7 81
maximum length

Manual phalanx III-1 length 9 il
Manual phalanx III-2 length 72.6 Bal
Manual phalanx III-3 length 6.7 TA
iste phalanx III-4 (ungual) 949 Gil

maximum length

Table 2 - Anatomical proportions of the holotype of Scipionyx samniticus (selected ratios).

Skull / Presacral 0.48
'Skull / Femur 1:99
Humerus / Femur 0.71
\Forelimb / Presacral (skull included) 0.48
Orbit / Skull 0.33
Humerus / Scapula 21 |

109

Pelvic girdle and hind limbs Right leer
Ilium length 26.7 -
Ilium preacetabular length 11.0 -
Ilium postacetabular length 9.4 -
Ilium height at pubic process 11.5 -
Ilium height i 59 g

at supracetabular margin i

Pubis length ZIA -
Pubis minimum shaft diameter 15) -
Pubic foot length 4.6 -
Ischium length -20.4 -
Ischium minimum shaft diameter 1.6 21.4
Ischial foot length ssi -
Femur length APE) -
Femur proximal diameter gia -
Femur minimum 48 x
mid-shaft diameter |

Femur distal diameter 6.4 77.0
Tibia preserved length 14.7 8.5
Tibia proximal diameter DIO 4.7
Tibia minimum preserved 1; È
Shaft diameter |

Fibula preserved length 14.5 4.7 |
Fibula proximal diameter 4.1 -
Fibula minimum preserved 1g _
shaft diameter i

Radius / Humerus 0.67
Manus / Forelimb 0.42
Length/width of McI-III 2.00
Length digit III / digit I (25
Pubis / Femur 0.73
Pubis / Ischium 1.34

110 CRISTIANO DAL SASSO & SIMONE MAGANUCO

ONTOGENETIC ASSESSMENT

Introduction - Although several individuals of em-
bryonic, hatchling and juvenile dinosaurs have been
found and described especially in recent years (e.g.,
Coombs, 1982; Horner & Currie, 1994; Norell et al.,
1994; Mateus et al., 1998; Carpenter, 1999; Norell et al.,
2001; Xu et al., 2001; Varricchio et al., 2002; Rauhut &
Fechner, 2005; Reisz et al., 2005; Goodwin et al., 2006;
Schwarz et al., 2007a; Balanoff et al., 2008; Kundrat
et al., 2008; Bever & Norell, 2009), the fossil record
of theropods representing early stages of skeletal ontog-
eny with well-preserved cranial material, remains ex-
tremely poor (e.g., Rauhut & Fechner, 2005; Bever &
Norell, 2009). Scipionyx is one of these few theropods
and is probably the best preserved one. It was identi-
fied as a juvenile individual by Dal Sasso & Signore
(1998a, 1998b) and as probably a hatchling by Holtz et
al. (2004), who pointed out the difficulty in determining
whether the generally plesiomorphic condition of this
dinosaur is due to a primitive position of the taxon in
the coelurosaur tree or to the early ontogenetic stage of
the type specimen. Therefore, recognition of the ontoge-
netic stage of Scipionyx samniticus is important for two
aspects: it is critical for proper anatomical comparison
and subsequent interpretation of character states in this
species (see Phylogenetic Analysis); and it may improve
our understanding of skeletal development and its phy-
logenetic patterns within Theropoda and, in the final
analysis, of the evolutionary history of this group.

In assessing the ontogenetic stage of Scipionyx, it
must be taken into account that ontogeny of a genus or
of a “family-level” taxon does not exist (Steyer, 2003).
Developmental events can occur at different times along
the ontogenetic trajectories of different lineages, even in
strictly related taxa. Therefore, a growth series within a
unique species, with intraspecific variations, is needed to
study its development.

Unfortunately, Scipionyx samniticus is known only
from a single specimen, and virtually no growth series
are known among its closest relatives, the Compsogna-
thidae. Juravenator, Huaxiagnathus and Sinocalliopteryx
are also known from single specimens. The only species
represented by more than one specimen are Compso-
gnathus and Sinosauropteryx. According to Peyer (2006),
who based her interpretation not merely on whole body
size, both specimens of Compsognathus longipes are ju-
veniles, with the German specimen — the holotype de-
seribed by Ostrom (1978) — being more immature than
the French specimen. Concerning Sinosauropteryx, few
specimens have been described to date: Currie & Chen
(2001) considered the small holotype NIGP 127586 a
Juvenile, and specimen NIGP 127587 a young adult —
more precisely, an animal of reproductive age and prob-
ably approaching mature size, but still young, when it
died; the specimen described more recently by Ji et al.
(2007b) is intermediate in size, but its poor preservation
renders it difficult to compare it with the other specimens
and to recognise whether it is also intermediate ontoge-
netically. Gò6hlich & Chiappe (2006) considered Jurave-
nator a Juvenile, based on features such as bone texture

and the degree of fusion of the neural arches. Hwang

et al. (2004) briefly commented on the ontogenetic as-
sessment of Huaxiagnathus, suggesting that it is juve-

nile (but see Incomplete Ossification Of The Vertebral
Column), whereas Sinocalliopteryx (Ji et al., 2007a),
which is the largest compsognathid known, seems to be
a mature animal.

Therefore, as mentioned above, a developmental se-
quence for any compsognathid species is still not known,
and it is not our intention to reconstruct one for the whole
Compsognathidae taxon based on these composite data,
as we consider this practice inadequate and unorthodox.
However, some allometric trends may be tentatively iden-
tified and discussed for Scipionyx, based on comparisons
with other compsognathids (considered in their aforemen-
tioned ontogenetic stages) and theropods, and sometimes
extending the comparison to other extinct and extant ar-
chosaurs and, more generally, vertebrates, with the aim of
assessing the ontogenetic stage of Scipionyx.

Dal Sasso & Signore (1998a) interpreted Scipionyx as
a juvenile, based on the following set of characters: gen-
eral body proportions; high skull/presacral length ratio;
short antorbital region; large, circular orbit; many unfused
skeletal elements (scapulocoracoids, sternal plates, sacral
vertebrae); several neural arches separated from their
centra; symmetry of tooth development as an indication
that the first replacement had not yet occurred; and low
denticle count. All these characters that potentially give
some indication of the ontogentic stage of the holotype of
Scipionyx samniticus, plus some new ones, are discussed |
in detail in the following paragraphs.

Gut contents - The first, unequivocal datum useful
for assessing the ontogenetic stage of Scipionyx is the
presence of gut contents (see Gut Contents And Feeding
Chronology): this clearly indicates that the individual is
not embryonic.

Scarred bone surfaces - Incomplete ossification of
the periosteum is particularly evident in the long bones
of sauropod embryos from the Morrison Formation, USA
(Britt & Nailor, 1994) and Auca Mahuevo, Patagonia (Sal-
gado et al., 2005), and in the embryonic theropods from
Lourinhà, Portugal (Mateus, pers. comm., 2005). In these
embryos, the periosteum exhibits the porous appearance
typical of embryonic and hatchling archosaurs (Bennett,
1993; Sanz et al., 1997; Horner, 2000; de Ricqlès et al.,
2000). Such a porous appearance, with furrows and pits
producing a ‘‘scarred” effect, tends to disappear gradually
during growth, as can be inferred from fossil specimens
and clearly seen in extant birds (Chiappe, pers. comm.,
2006). The scarred effect is marked on the long bones of
the juvenile theropods, such as Sinornithoides (Currie &
Dong, 2001) and Juravenator (Gòhlich & Chiappe, 2006),
in immature extant crocodilians (Dal Sasso & Maganuco,
pers. obs., 2010, on a Crocodvylus niloticus hatchling; Fig.
108A) and birds, and in the young adult Sinosauropteryx
(Currie & Chen, 2001). A striated bony texture character-
ises also the maxilla from Guimarota (Portugal) referred
to an early post-hatching individual by Rauhut & Fech-
ner (2005). The extensive “scarred” texture of the bone
surfaces of Scipionyx (e.g., on the external surface of the
lower jaw, scapular blade and femural shaft; Fig. 108B-D)
closely resembles that of embryos and hatchlings, rather
than that of late juvenile or young adult individuals. Al-

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 111

Fig. 108 - The bone surface of a Crocodylus niloticus hatchling
(A, femur) is comparable in texture with that of Scipionyx samniticus
(B, dentary; C, scapula; D, femur). Scale bar = 1 mm.

Fig. 108 - Tessitura ossea fittamente solcata sul femore di un Cro-
codylus niloticus neonato (A), messa a confronto con la tessitura, del
tutto analoga, presente sulle ossa di Scipionyx samniticus (B, dentale;
C, scapola; D, femore). Scala metrica = 1 mm.

so, this scarred aspect does not look in any way like that
seen as a consequence of exaggerated bone reabsorption
in older, stressed crocodiles (Huchzermeyer, 2003: fig.
6.10) or like the well-developed pits and grooves form-
ing articular rugosities in the articular surfaces of mature
specimens of the sauropod Camarasaurus (Schwarz et al.,
2007b). Thus, the intensively scarred bone surface clearly
indicates that Scipionyx is not a mature animal, and that it
is more likely a hatchling.

Large skull - The skull of Scipionyx is very large with
respect to its body. The skull/presacral length ratio (0.48)
is higher than in the vast majority of the well-known adult
theropods (e.g., Weishampel e? a/., 2004) but comparable
to that of the tyrannosaurids, some compsognathids (see
below) and some dromaeosaurids (e.g., Ostrom, 1969;
Burnham ef a/., 2000). The carcharodontosaurids, the re-

cently found long-snouted tyrannosauroids (e.g., Li et al.,
2009) and the gondwanan dromaeosaurids (e.g., Novas et
al., 2008b) are too incomplete in the postcranium to per-
mit any comparison. A negative growth allometry of the
skull respect to the postcranium is common in most ex-
tant and extinct vertebrates: compared with the rest of the
body, the juvenile skull is relatively larger than the adult
one. A similar estimation can be obtained by comparing
the length of the skull to that of the femur. In the holotype
of Compsognathus, the skull length is 72 mm, represent-
ing some 40% of the length of the presacral vertebral col-
umn. In the French Compsognathus, the skull is long and
slender and has a reconstructed maximal length of 100
mm, i.e., about 30% of the length of the presacral ver-
tebral column (as in Peyer [2004]; contra Peyer [2006],
who reported that the skull is 22% of the length of the
presacral vertebral column but, according to figures and
measurements, it is in reality 22% of the whole presacral
length, skull included). The skull/presacral length ratio in
Sinosauropteryx ranges from 0.5 in the smallest individ-
ual to 0.35 in the other two, although in the third, poorly
preserved specimen (Ji et a/., 2007b), this ratio cannot be
measured with precision based on the published photo-
graphs. The remaining well-known compsognathids are
represented by a single specimen each, all of them hav-
ing a large skull, with a ratio possibly greater than 0.5
in Juravenator, approaching 0.5 in Huaxiagnathus and
greater than 0.4 in Sinocalliopteryx.

Summing up, the Compsognathidae have a proportion-
ally large skull in early/late juveniles or adults. So, based
on the available data, it is not clear whether the compara-
tively large size of the skull of Scipionyx can be interpreted
as a Juvenile trait or a representative feature which would
have been maintained, with little or no modification, dur-
ing ontogeny, as seems to occur in some compsognathid
species such as — at least — Sinocalliopteryx gigas.

Therefore, juvenile features must be ascribed to the
skull based on the shape and the proportions of its vari-
ous elements/portions, rather than on overall skull size
with respect to the rest of the body. A similar conclusion
can be drawn for the hyoids, which seem to be oversized
in Scipionyx at first glance (Holtz, pers. comm., 1998).
The skull/hyoid length ratio is 2.6 in Scipionyx, about 3.6
in Huaxiagnathus and 3.7 in Sinosauropteryx (Ji et al.,
2007b). A close value (i.e., 2.5) is found in the French
Compsognathus (Peyer, 2006), which has a head that is
comparatively smaller than that of Scipionyx. Large hy-
oids are present also in Proceratosaurus, in which the
ratio is 2.2 (Rauhut et al., 2009). A ratio of 4.2 can be
estimated for Bambiraptor (Burnham et al., 2000). Based
on those data it appears that, as expected, the size of the
hyoids parallels that of the head.

Orbital foramina rounded and proportionally very
large and antorbital region short and deep - In juvenile
dinosaurs, as in the vast majority of extant and extinct ver-
tebrates, the snout is relatively short and deep (e.g., Norell
et al., 1994), and the orbits are proportionally large with
respect to the length of the skull (e.g., Coombs, 1982; Car-
penter, 1994; Horner & Currie, 1994; Weishampel et al.,
2004; Rauhut & Fechner, 2005; Reisz et a/., 2005; Tykoski,
2005: Goodwin et al., 2006; Kundrat et al., 2008). These
characters are size- and age-related, and are subjected to
strong allometry in relation to other portions of the skull

N22 CRISTIANO DAL SASSO & SIMONE MAGANUCO

during growth (e.g., Currie & Dong, 2001). In particular,
the orbits are subjected to negative allometry, as they tend
to become relatively smaller. They reach their maximum
relative size in embryos, as examplified by the titanosaur
embryos preserving the skull (Salgado er a/., 2005), in
which the orbits are at least 50% of the skull length.

The orbits are relatively large and rounded in shape
in species represented by immature individuals, such as
Compsognathus longipes and Juravenator starki, but
also in many adult theropods which attained small-to-
medium body size in adulthood, independently of their
phylogenetic affinities (e.g., Currie, 2003; Weishampel et
al., 2004). This can be seen in the juvenile holotype and
in the young adult individual of Sinosauropteryx prima
described by Currie & Chen (2001): although both suf-
fered post mortem dorsoventral crushing, it is interesting
to note that although the largest specimen has a larger
skull, the orbits were reconstructed with approximately
the same shape and absolute size. The orbits are subjected
to a stronger negative allometry in the adults of species
that attained a large body size (e.g., tyrannosaurids), in
which they also become keyhole shaped.

In small- to medium-sized coelurosaurs, the antorbital
portion of the skull tends to increase in length much more
than the orbital and postorbital portions, showing a strong
positive allometry. As a consequence, both orbits and fron-
tals appear larger in juveniles. The orbits of Scipionyx are
intermediate in size between those of the embryos and oth-
er juveniles. In Scipionyx, the orbit is twice as long as the
antorbital fenestra (Fig. 26), just like in the Juvenile speci-
mens IVPP-V11797-10 and IVPP-V11797-11 of Sinorni-
thomimus (Kobayashi & Lii, 2003) and in the perinate By-
ronosaurus IGM (= MGI1) 100/972 (Bever & Norell, 2009),
in which the snout looks even shorter than in Scipionyx.
In a therizinosauroid embryo (Kundrat et al., 2008), the
length of the orbit is at least three times that of the antor-
bital fenestra, which is higher than it is long (longer than
high in adult Erlitosaurus). In juvenile individuals, such as
the dromaeosaur Bambiraptor (specimen FIP 001 [Burn-
ham et al., 2000]) and the compsognathids Huaxiagnathus,
Juravenator, Sinosauropteryx and Compsognathus, the
antorbital fenestra approaches in length that of the orbit,
whereas it is equal or even longer in the putative mature
holotype of Sinocalliopteryx. In the Guimarota hatchling
referred to A/losaurus (specimen IPFUB Gui Th 4 [Rauhut
& Fechner, 2005]), only the maxilla is preserved and, simi-
lar to Scipionyx, it 1s short and high, with an almost vertical
rostral margin of the antorbital fenestra and a low maxil-
lary body. Certainly, Scipionyx and most of the other thero-
pods, including non-coelurosaur tetanurans (e.g., Rauhut
& Fechner, 2005) and non-tetanuran taxa (e.g., Tykoski,
2005), show the opposite condition to that of the tyranno-
saurid coelurosaurs, in which the snout is already elongate
in uveniles, and there is little ontogenetic change in the rel-
ative length of the antorbital fenestra during growth (Cur-
rie, 2003). Thus, based on these data, the relative size of the
orbit in relation to the antorbital fenestra indicates that the
Scipionyx specimen is a hatchling or an early juvenile.

Position of the maxillary fenestra - In lateral view,
the maxillary fenestra of Scipionyx is located more dor-
sally and more rostrally than in the other compsognathids,
and its rostral margin coincides with that of the antorbital
fossa (Fig. 26). As a consequence, the maxillary fenestra

seems to be bordered by the maxillary medial wall of the
antorbital fossa only caudally and ventrally. In the other
compsognathids, the maxillary fenestra opens within the
maxillary medial wall of the antorbital fossa, so that part
of this wall is visible rostrally to the rostral margin of the
maxillary fenestra in lateral view, separating the latter
from the rostral margin of the antorbital fossa. The condi-
tion in Scipionyx might be related to the immaturity of the
specimen and/or to the consequent shortness of its snout.

An opposite trend can be seen in the ontogenetic se-
ries of Tyrannosaurus (Carr & Williamson, 2004: fig. 6)
and in other tyrannosaurids (e.g., Currie & Dong, 2001):
the maxillary fenestra, which appears within the maxil-
lary medial wall in juveniles, migrates towards the rostral
margin of the antorbital fossa in adults, in parallel with
the growth of the snout, which becomes bulkier.

Rostral ramus of the maxilla - In Scipionyx, the ros-
tral ramus of the maxilla is little developed and is shorter
than high (Fig. 29). In Juravenator and in the holotype of
Sinosauropteryx, it is slightly longer than high. In Hua-
xiagnathus and Sinocalliopteryx, it is almost twice as long
as high, and in the French Compsognathus (Peyer, 2006)
it is intermediate in size. Although no ontogenetic series
is known for compsognathids, this portion of the maxilla
undergoes, in all likelihood, great elongation during on-
togeny (see also Maxilla).

Unfused interdental plates - Contrary to Dal Sasso &
Signore (1998a), unfused interdental plates (Fig. 30) are
not necessarily indicative of the juvenile stage of Scipio-
nyx. As a matter of fact, they are unfused also in many
other compsognathids (see Maxilla) and do not fuse even
in the late adult of tyrannosaurid coelurosaurs (e.g., Cur-
rie et al., 2003).

Nasals shorter than frontals - The nasals in Comp-
sognathus (Peyer, 2006) and Juravenator (Gòhlich &
Chiappe, 2006) are proportionally much longer than in
Scipionyx and make up almost half of the length of the
skull. They are even longer in Sinocalliopteryx. In adult
theropods, the nasals are shorter than the frontals only in
oviraptorosaurs (Rauhut, 2003). In the perinatal skull of
Byronosaurus (Bever & Norell, 2009), the nasal is com-
parable in length to the frontal. Unfortunately, the fron-
tal is not preserved in more mature individuals (Norell
et al., 2000). The nasal/frontal length ratio is approxi-
mately 1/3 in the therizinosauroid embryo and 1/1 in the
adult Erlikosaurus (Kundrat et al., 2008). The presence
of shorter-than-frontal nasals in Scipionyx (Figs. 24, 25A))
supports its immaturity and suggests that, together with
the maxilla, the nasals undergo positive allometric growth
associated with a rostral elongation of the snout region
during postnatal ontogeny. A similar pattern of positive
allometric growth of the nasals has been reported in the
growth series of the skull of the prosauropod dinosaur
Massospondvlus (Reisz et al., 2005).

Position of the caudal margin of the apertura nasi
ossea - In Scipionyx, the caudalmost portion of the naris
terminates slightly caudally to the rostralmost border of
the antorbital fossa (Fig. 26; see Apertura Nasi Ossea).
Scipionyx is the only compsognathid with such a partial,
virtual overlapping, although a similar condition occurs

SCIPIONYX SAMNI / } J Ù
YX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 113

in Huaxiagnathus, where the caudal margin of the naris
seems to reach the level of the rostral margin of the fossa.
According to Chiappe (pers. comm., 2006), the condition
in Scipionyx might be related to the immaturity of the
specimen and to the shortness of its snout, in particular
of the rostral ramus of the maxilla and of the nasals. This,
however, cannot be definitely demonstrated, as the same
overlapping occurs in the basal tyrannosauroids Proce-
ratosaurus (Woodward, 1910) and Guanlong (Xu et al.,
2006), and in other more derived coelurosaurs such as /n-
cisivosaurus (Xu et al., 2002a), Sinovenator (Xu et al.,
2002b) and Archaeopteryx (Elzanowski, 2002); moreover,
the two margins are very close in Di/ong, Ornitholestes
(Osborn, 1916: fig. 1; Rauhut, 2003: fig. 5H) and, poten-
tially, in the already mentioned Huaxiagnathus.

Shape of the lacrimal - In Scipionyx, the horizontal
ramus of the lacrimal is shorter than the vertical ramus
(Fig. 25A). In Juravenator, which is a juvenile specimen,
both rami are subequal in length. In the French Comp-
sognathus, a late juvenile, and in Sinocalliopteryx, an in-
dividual close to maturity, the horizontal ramus is longer
than the vertical ramus. The condition in Scipionyx likely
reflects its earlier ontogenetic stage.

Large prefrontals, with descending lateral process
well-exposed in lateral view - In Scipionyx, the prefron-
tal is large and robust, is rostrocaudally longer than the
horizontal ramus of the lacrimal, and has a descending
lateral process that is well-exposed in lateral view (Fig.
25A). Brochu (2003) proposed that in tyrannosaurids the
prefrontals are ontogenetically ephemeral structures that
gradually merge with the lacrimals. In perinates Byrono-
saurus, there is no evidence of separate prefrontal ossi-
fication, as described for in other troodontids (Bever &
Norell, 2009). The prefrontal is absent in Archaeopteryx,
too, and in all other birds (e.g., Elzanowski, 2002). In con-
trast, the prefrontal is present and visible in lateral view
in a large number of coelurosaurs, including compsog-
nathids, basal tyrannosauroids, therizinosauroids (where
it is comparatively larger in the embryos [Kundrat ef
al., 2008]), alvarezsaurids and ornithomimosaurs, being
particularly large in some ornithomimosaur species (for
further comparisons, see Prefrontal). Thus, among coe-
lurosaurs, Scipionyx is unusual particularly in the ventral
prolongation of the bone and in its comparatively large
size, although it is not as large as in the ornithomimosaurs.
Based on the present data, it is impossible to establish to
what extent the comparatively large size and exposition
of this bone in Scipionyx is related to its early ontogenetic
stage (i.e., earlier than that of any other compsognathid),
or if this is a characteristic of the taxon that would have
persisted in adults, or partly both.

Sublacrimal expansion of the jugal - The horizontal
body of the jugal of Scipionyx lacks any sublacrimal ex-
pansion (Fig. 25A). The same occurs in other juveniles,
such as Compsognathus (Peyer, 2006) and Juravenator
(Gohlich & Chiappe, 2006). On the other hand, the pos-
sibly early adult Huaxiagnathus (Hwang et al., 2004: 2a)
shows a feeble expansion, whereas the mature Sinocal-
liopteryx has a distinct expansion (Ji ef a/., 2007a: Fig.
2b). The condition in the immature individuals might re-
flect their earlier ontogenetic stage.

DD

Frontoparietal fontanelle open - The skull roof of
Scipionyx unequivocally has a frontoparietal fontanelle
(see Frontal), which is well-open (Figs. 23-34). As re-
ported by Balanoff & Rowe (2007), the frontoparietal
fontanelle is a diamond-shaped space between the paired
frontals and parietals, the size of which is an indica-
tor of age (i.e., the larger the opening, the younger the
specimen). In extant crocodilians, there are no signs of
a frontoparietal fontanelle after hatching (Kundrat ef
al., 2008; Brochu, pers comm., 2009). The same occurs
in the Triassic sauropodomorph Mussaurus, in which
there are no signs of a frontoparietal fontanelle in ei-
ther post-hatching or subadult specimens (Pol & Pow-
ell, 2007). According to Salgado ef a/. (2005), there is
a frontoparietal fontanelle in titanosaur embryos from
the Upper Cretaceous of Patagonia. The existence of a
frontoparietal fontanelle has been proposed in Diplodo-
cus, Camarasaurus, Dicraeosaurus and Amargasaurus
(Salgado et al., 2005; and reference therein). In the last
two genera, the fenestra remains open in adults, whereas
in the remaining taxa it is apparently present only in im-
mature individuals.

Unfortunately, few data are available for Mesozoic
theropods. A frontoparietal fontanelle has been identified
in the perinates referred to the troodontid Byronosaurus
(Bever & Norell 2009). In contrast, no signs of an open-
ing were recognised on the caudomedial edge of the fron-
tal in a therizinosauroid embryo (Kundrat er a/., 2008).
No signs of the frontoparietal fontanelle have been re-
ported either in the other immature theropods, including
composognathids, in all likelihood because none of these
specimens is of a hatchling and at the age of death, the
fontanelle, even if present when hatching, was probably
already closed.

More data are available for extant avian theropods. In
the zebra finch, Taeniopygia castanotis, which leaves the
nest 21 days after hatching, the frontoparietal fontanelle
greatly decreases in size between days 24-27 and finally
disappears after 36 days (Serventy et a/., 1966). The same
trend can be seen in the Indian weaver bird, Ploceus phi-
lippinus, which leaves the nest 17-18 days after hatching,
and has the complete closure of the fontanelle around day
40 (Biur & Thapliyal, 1972). In the ratites Struthio and
Rhea, the fontanelle does not close until well-after hatch-
ing, and is still well-visible for 10 days (Balanoff., pers.
comm., 2009).

Based on these composite data, we can suppose that
the frontoparietal fontanelle remained open in hatchling
coelurosaurs, disappearing a few weeks later, just like
in some extant forms. If this postnatal trajectory is con-
firmed within coelurosaurs or more inclusive nodes, then
the presence of a comparatively large frontoparietal fonta-
nelle in Scipionyx indicates that it could have had no more
than two weeks when it died.

Elements of the braincase in loose contact - The
visible elements of the braincase of Scipionyx have loose
contacts (Fig. 25C). Among the examples of uncontrover-
sial ontogeny-related characters present in most dinosaurs
is co-ossification and fusion of braincase elements (e.g.,
Tykoski, 2005). The condition of Scipionyx is the one ex-
pected in a juvenile individual; nevertheless, this does not
help to clarify its degree of maturity because the fusion of
these elements occurs at different times along the ontoge-

114 CRISTIANO DAL SASSO & SIMONE MAGANUCO

netic development of different theropod taxa, and often
close to maturity; moreover, some interorbital braincase
elements do not ossify at all by adulthood in some line-
ages (e.g., Carrano et al., 2002).

Vomers unfused to each other all along their length -
Madsen (1976) reported that “evidently the vomers are
not paired in A//osaurus fragilis, but, if they are, fusion
must occur embryonically because even the smallest, ap-
parently juvenile, examples show no separation”. The
vomers are indeed fused for nearly all their length in many
adult theropods (e.g., Clark ef a/., 2002; Currie, 2003;
Holtz et al., 2004). The condition in Scipionyx in all like-
lihood reflects its early ontogenetic stage (Figs. 25C-D,
37). It must be taken into account, however, that there are
some exceptions among adult individuals of coelurosaurs,
such as /Incisivosaurus, in which the vomers are unfused
along most of their length and are fused only dorsally at
the rostral end of the element (Balanoff e? a/., 2009), and
Velociraptor, which is reported to be unique among thero-
pods in having the vomers distinctly separated along their
entire length (Barsbold & Osmolska, 1999).

Craniocaudal length of the dentary - In Comp-
sognathus (Ostrom, 1978; Peyer, 2006), Huaxiagnathus
(Hwang et al., 2004), Sinosauropteryx (Currie & Chen,
2001) and Juravenator (Gòhlich & Chiappe, 2006), the
dentary is proportionally longer than in Scipionyx. This
suggests that during ontogeny this bone undergoes greater
elongation than the rest of the lower jaw. As for the snout,
its relative shortness is another indication of the immatu-
rity of the specimen.

Angular moved dorsally - As mentioned in the de-
scription of this bone, during diagenesis the right angular
lost its contact with the dentary and has moved dorsally,
partially overlapping the surangular (Fig. 25B). It is not
possible to establish whether the weak sutural contact
between the angular and the dentary is primarily linked
to the ontogenetic stage of Scipionyx or to the presence
of an intramandibular joint between the dentary and the
postdentary bones, as is the case in most basal tetanurans
(Holtz et al., 2004).

Symmetrical tooth crown height in counterlateral
tooth rows - The symmetrical development of the tooth
series in the left and right tooth rows of Scipionyx (Figs.
24, 44-46) is consistent with tooth replacement having
not yet started. As a matter of fact, there are no alveoli
with erupting replacement teeth, with the possible ex-
ception of the right m3, which is slightly smaller than
the left m3. Empty alveoli were clearly identified, for
example, by Ostrom (1978) in the juvenile holotype of
Compsognathus.

In archosaurs, tooth formation starts in embryos, and
the first wave of teeth erupts before hatching. This has
been observed, for example, in extant crocodilians (Mar-
tin, pers. comm., 2009), fossil theropods from Lourinhà,
Portugal (Dal Sasso & Maganuci, pers. obs., 2005) and
fossil titanosaurs from Patagonia (Salgado et al., 2005).
First replacement occurs in just weeks in hatchling croc-

odilians, and probably the same occurred in hatchling

dinosaurs (Erickson, pers. comm., 2009). In Alligator
mississipiensis, several generations of teeth are already

replaced during embryonal development and, after hatch-
ing, the first teeth of the functional group are expected to
be shed after 3-4 weeks (Westergaard & Ferguson, 1990;
Huchzermeyer, pers. comm., 2010). In archosaurs, tooth
replacement rate seems to be faster in juveniles than in
adults (Erickson, 1996a; D’Emic, pers. comm., 2009),
possibly as a direct consequence of the very fast growth
rate ofthe tooth-bearing bones at this developmental stage
(Martin, pers. comm., 2009).

Concerning the tooth growth rate, some information
can be inferred by the incremental lines of von Ebner: these
are microstructural features that demarcate the daily appo-
sition of dentin, such that the total number of incremental
lines in a tooth serves as a measure of its age (Erickson,
1996a, 1996b; D’Emic, 2009). Given the uniqueness and
the aesthetic value of the Italian compsognathid, we re-
fused to thin-section a tooth of Scipionyx in order to count
the incremental lines of von Ebner. However, because the
incremental lines do not vary much in thickness in archo-
saur taxa (Erickson, 1996b), being around 10-20 um each
in most teeth, tooth-formation times are expected to be
in the order of weeks in teeth such as those of Scipionyx
(D’Emic, pers. comm., 2009).

Summing up, the fact that tooth replacement had not
yet started, and the dimension of Scipionyx £ teeth both
suggest that this individual was not more than a few
weeks old at death.

Low number of lateral teeth - The dental formula
in Scipionyx is Spm+7m/10d teeth, whereas most non-
avian theropods bear at least 11-12 teeth in the maxilla
(Tykoski, 2005). The low number of lateral teeth in Sci-
pionyx might either reflect a low number of teeth in this
species or is related to ontogeny, as a consequence of the
rostrocaudal shortness of the lateral tooth-bearing bones
(the dentary and the maxilla). As these bones have been
recognised among the elements which generally undergo
greater elongation during ontogeny, the formation of new
tooth positions would be expected during growth if the
teeth maintain the same size relative to the dorsoventral
depth of the bone. Concerning the maxilla, the formation
of new tooth positions seems to occur in Compsognathus
longipes: there are 17-18 maxillary teeth in the more ma-
ture French specimen, and 15-16 in the more immature
German specimen, suggesting that this intraspecific varia-
tion might be related to ontogeny (Peyer, 2006). Rauhut &
Fechner (2005) noted an increase in the number of maxil-
lary teeth, from 13 to 16 (more rarely 15), in A//osaurus
fragilis during postnatal ontogeny. Bever & Norell (2009)
offered three possible interpretations of the small number
(13-15) of maxillary teeth for a troodontid in the Mon-
golian perinates: it represents the retention of the plesio-
morphic paravian condition (based on comparison with
dromaeosaurs); it represents a secondarily derived condi-
tion within troodontids (this implies that the perinates be-
long to a new species of Byronosaurus);, or Byronosaurus
exhibits a significant postnatal increase in the number of
maxillary teeth, without a significant relative increase in
the length of the maxillary tooth row.

The number of maxillary teeth may increase through
ontogeny also in basal sauropodomorphs (Galton, 1990)
and ornithischians (Varricchio, 1997), whereas an op-
posite trend occurs in tyrannosaurids, in which the teeth
undergo great changes during ontogeny: an increase in

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

tooth width is accompanied by a loss of 3-5 tooth posi-
tions (Carr, 1999). Based on Varricchio (1997) and Carr
(1999), Rauhut & Fechner (2005) considered the increase
in the number of maxillary teeth to be the plesiomorphic
postnatal trajectory within Tetanurae, with coelurosaurs
exhibiting a derived trajectory in which the number of
maxillary teeth during postnatal growth either remained
stable or decreased. Compsognathus and Byronosaurus,
however, seem to indicate that the scenario is more com-
plex within Coelurosauria.

Concerning the fact that most non-avian theropods
bear at least 11-12 teeth in the maxilla, 12 maxillary teeth
are indeed reported in the NIGP 127587 specimen of Si-
nosauropteryx (Currie & Chen, 2001). Only 6 maxillary
teeth have been reported in Sinocalliopteryx, although at
least 12 tooth positions seem to be present based on the
right maxilla in medial view (Ji et al., 2007a: fig. 2a). If
this interpretation is correct, in that specimen there is an
alternation of empty and occupied alveoli, reflecting the
pattern visible in the lower jaw, which bears short and tall
crowns. Ostrom (1978) interpreted the similar alternation
of empty and occupied alveoli in Compsognathus as a di-
rect evidence of the alternate pattern of the replacement of
the teeth. A low number of maxillary teeth is reported for
the remnant compsognathids: 8 maxillary teeth in Jurave-
nator (GGhlich & Chiappe, 2006) and at least 8 maxillary
teeth in Huaxiagnathus, although the tooth count in its
maxilla is not clear (Hwang et al., 2004). Comparatively
few maxillary teeth are known also in other, not strictly
related, small theropods, such as /ncisivosaurus (9-Ba-
lanoff e? al., 2009), Bambiraptor (9-Burnham et al.,
2000), Archaeopteryx (8/9-Elzanowski, 2002) and Noa-
saurus and Masiakasaurus (10 or fewer-Tykoski & Rowe,
2004; Tykoski, 2005).

Summing up, based on all these data, we conclude that
the maxillary tooth count may be affected not only by the
ontogenetic stage of the individual but also by the overall
body size attained by the species. This difficulty in the in-
terpretation of the data and in the recognition of postnatal
trajectories reflects both the scarcity of ontogenetic data
and the complexity of theropod evolution.

Teeth with few denticles - In Scipionyx, denticles
are present on the distal carinae of lateral teeth for 3/5 to
4/5 of the crown height in crowns not exceeding 3 mm in
height; the denticle count is approximately 13 denticles
per mm (Figs. 45-48). In the perinate theropods referred
to Byronosaurus (Bever & Norell 2009), no serrations are
present, as in the adult individuals of that taxon and of
other troodontids. A comparison can be made with the
ML565 specimen — a right maxilla of a theropod embryo
from Lourinhà (Mateus er a/., 1998) — in which the best
exposed maxillary crown measures no more than 1 mm in
height, the basal half of the distal carina bears 4-5 well-
developed denticles and the denticles gradually tend to
disappear in the apical half (Mateus, pers. comm., 2004;
Dal Sasso & Maganuco, pers. obs., 2004). In some thero-
pods, the size of the denticles decreases at both apical and
basal ends of the carinae, but even when this occurs, the
denticles do not completely disappear in the apical half
of the carina (e.g., Maganuco et a/., 2005; and references
therein). Denticles that are well-developed only in the ba-
sal half of the carina might reflect the ontogenetic stage of
the embryos from Lourinhà. The condition in Scipionyx

115

might be intermediate between that of these embryos and
that observed in non-neonate juveniles and adult individ-
uals of other species. More data are required in order to
support or challenge this preliminary hypothesis.

Denticle size and density - Within a single theropod
species, the absolute size of the denticles is related to tooth
size, which in turn is related to skull size and body size,
without any particular variation during ontogeny. In other
terms, the teeth of the small juvenile individuals look like
a scaled-down version ofthe teeth ofthe large adults ofthe
same species. As a consequence, within a species the den-
sity per mm is higher in smaller individuals because they
have smaller teeth. In all likelihood, therefore, the very
high density of denticles in Scipionyx is related mostly to
the absolute small size of the specimen, and does not help
in the assessment of its ontogenetic stage. Among comp-
sognathids, in fact, Sinosauropteryx has 11-14 denticles
per mm, Compsognathus has no more than 12 denticles
per serrated margin in crowns up to 4 mm high — although
it must be taken into consideration that they are present
only in the upper third of the distal carina where they are
about 9 per mm — Huaxiagnathus has 7 denticles per mm
and Sinocalliopteryx, which is the largest known genus,
has about 4 denticles per mm.

Incomplete ossification of the vertebral column - In
the postcranial skeleton, evidence that an archosaur was
neither fully mature nor fully grown at death is provided
by the incomplete ossification of the vertebral column
(e.g., Currie & Zhao, 1993a; Brochu, 1996; Xu et al.,
2001; O’Connor, 2007; Schwarz ef al., 2007a). The pat-
tern of closure of the neurocentral suture within Archo-
sauria is not uniform. Kobayashi & Li (2003) pointed out
that in the holotype of Sinornithomimus, which is con-
sidered a juvenile individual, the neurocentral sutures are
fused in all of the cervical vertebrae and in the first three
dorsal vertebrae, whereas the rest are unfused. In the sub-
adult Majungasaurus specimen described by O°Connor
(2007), the postaxial cervical neural arches are fused to
their respective centra, the first two dorsals are partially
fused, the remainder of the dorsal vertebrae are unfused,
the preserved sacrals are fused, the other sacral compo-
nents are not fused, as are neither some proximal caudals.
Not completely fused neurocentral sutures are found al-
so in the dorsal and caudal vertebrae of the holotype of
Garudimimus brevipes (Kobayashi & Barsbold, 2005),
whereas complete obliteration of the neurocentral suture
can be seen in the allegedly super-precocial therizinosau-
roid embryo (Kundrat et al., 2008). These examples — but
many others are reported in the literature — suggest that
in non-avian theropods the closure of the neurocentral
sutures proceeds from the cervical vertebrae in a caudal
direction, with the sacrum preceding most of the dor-
sals, as in extant chelonians, squamates (Rieppel, 1992a;
1992b; 1993) and birds (e.g., Norell et a/., 1994; Balanoff
& Rowe, 2007). In contrast, in extant crocodilians closure
of the neurocentral sutures follows a distinct caudal-to-
cranial sequence (Brochu, 1996), with most of the cau-
dal vertebrae showing fully closed sutures in hatchlings,
even if the ossification sequence is craniocaudal as in the
other extant diapisds. A caudocranial direction in the clo-
sure of the neurocentral sutures seems to be present also
in sauropodomorph dinosaurs (Schwarz ef al., 2007a; and

116 CRISTIANO DAL SASSO & SIMONE MAGANUCO

references therein; see also Respiratory Physiology, this
volume).

Whatever the direction of the sequence of closure of
the neurocentral sutures, the process in Scipionyx was not
yet started, as is the case in one oviraptorid embryo ap-
parently close to hatching (Norell e? al., 1994; Norell et
al., 2001) and in one 7roodon embryo (Varricchio et al.,
2002). In Scipionyx, many vertebrae of which are disar-
ticulated and expose their neurocentral articular surfaces
(i.e., C2-3, C5, C7-10; DI, D4-11; S1, $4-5; Cal-3 and
Ca5-6: Fig. 109), all neural arches are unfused to their

centra (Figs. 50, 56, 64, 67). In the few cases in which the
arches are adjacent to their centra, the neurocentral su-
tures are well-marked. The neural arches preserved clos-
est to their respective centra are those of the cervicals, and
this condition is somewhat reminiscent of the sequence
of closure of the neurocentral suture seen in other thero-
pods. Over the length of the sacrum, all the visible axial
elements are disarticulated (neural arches, neural spines,
Îransverse processes and sacral ribs), whereas in mature
archosaurs these elements coalesce and fuse to the medial
walls of the ilia.

Fig. 109 - Line drawings highlighting the incomplete ossification of the vertebral column of Scipionyx samniticus. The red shading
identifies the exposed neurocentral articular surfaces. See Appendix 1 or cover flaps for abbreviations.

Fig. 109 - Incompleta ossificazione della colonna vertebrale in Scipionyx samniticus. Il reticolo rosso evidenzia le superfici articolari
neurocentrali esposte. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY I} er:

Among the other compsognathids, the French Comp-
sognathus has sacrals that are tightly co-ossified, except
for the last vertebra, i.e. the caudosacral (Peyer, 2006).
According to Currie & Chen (2001), some evidence of
immaturity can be seen in the axial skeleton of the holo-
type of Sinosauropteryx (specimen NIGP 127586). In
this specimen, most of the centra and neural arches are
not fused and a large attachment scar is present between
the sacral centra for a sacral rib, which was apparently
not fused to the sacral vertebrae at the time of death (a
scar for an unfused sacral rib has been reported also in a
Juvenile dromaeosaurid by Norell & Makovicky [1997]).
Gohlich & Chiappe (2006) reported the lack of fusion of
sacral vertebrae and the presence of open neurocentral
sutures in Juravenator. Hwang et al. (2004) supposed
that the holotypic and only known individual of Hua-
xiagnathus may represent a juvenile, based on the un-
fused neural arches in the proximal caudals and the large
size of the skull relative to the rest of the body. However,
a large skull seems to be a feature common to most comp-
sognathids (see Large Skull), and, from the published fig-
ures (Hwang et al., 2004: fig. 4), the proximal caudals
seem to be the only vertebrae in which the neural arches
are detached from their centra. As the caudal vertebrae
are the last to fuse in theropods, and the specimen also
has well-ossified carpals and tarsals, Huaxiagnathus may
have been a juvenile closer to maturity than previously
considered. Lastly, given the above-mentioned pattern of
closure of the neurocentral suture in theropods, Sinorni-
thoides youngi may be actually an immature individual
approaching maturity, as suggested by Currie & Dong
(2001), based on the fact that, among other things, the
neural arches are indistinguishably fused to the centra in
the middle and distal regions of the tail.

Cervical ribs not fused to the corresponding ver-
tebrae - The cervical ribs of Scipionyx are not fused to
the corresponding vertebrae (Fig. 73). Lack of co-fusion
of these elements has been considered as an indicator of
immaturity (e.g., Currie & Dong, 2001). Although the
cervical ribs are fused to their corresponding vertebrae in
some adult coelurosaurs (e.g., Norell et a/., 2001), Tyko-
ski (2005) reported that this is not the case in adults of
most basal tetanuran taxa; the character is also widely dis-
tributed among non tetanurans. In those taxa that exhibit
fusion in the adult, not all the cervical ribs fuse to their
vertebrae. Moreover, the cervical ribs are fused with the
corresponding vertebrae in the therizinosauroid embryo
(Kundrat ef a/., 2008), which was certainly immature. So,
it is apparent that this character is not strictly controlled
by ontogeny.

Regarding Scipionyx, we can conclude that the con-
dition exhibited in the specimen is the one expected in
immature individuals; however, although this cannot be
used to prove the ontogenetic assessment hypothesised
on the basis of the other data, it does not disprove it ei-
ther.

Unfused girdle elements - In Scipionyx, the girdle el-
ements are clearly unfused, with clearly visible lines of
separation (Figs. 87-88, 97-98, 103). The coracoids are
not fused to the scapulae in Sinosauropteryx (Currie &
Chen, 2001), Huaxiagnathus (Hwang et al., 2004), em-
bryos of therizinosauroids (Kundràt et al., 2008) and in

Juvenile coelophysoids (Tykoski & Rowe, 2004); howev-
er, scapula and coracoid are clearly fused in adults of the
latter group and of many other ones (e.g., Weishampel ef
al., 2004). The same lack of fusion characterises the ends
of the paired pubes in many taxa (e.g., Currie & Dong,
2001). Therefore, unfused girdle elements are often re-
lated to immaturity of the individuals. To complicate the
scenario, however, it must be taken into account that the
production of hormones (relaxin or relaxin-like) in repro-
ductive females of various vertebrates affects the com-
plete closure of some sutural contacts, such as the pubic
symphysis, which remains elastic to permit passage of the
eggs/foetuses. Similarly, it may have affected the proc-
ess in dinosaurs, as hypothesised for the pubic symphysis
of some adult A//osaurus specimens (Madsen, 1976; and
reference therein).

Relative size of the girdle elements - Girdles are
comparatively small in size in immature theropods such
as the juveniles of A/bertosaurus and the hatchlings of
Lourinhanosaurus (Mateus, pers. comm., 2004): as for
the ilium, this bone appears smaller, and craniocaudally
shorter compared to total body length, than what is ex-
pected in adults (Mateus, pers. comm., 2004). A com-
paratively short ilium is observed in Scipionyx (Figs. 21-
22). A similar condition is visible in Sinosauropteryx, in
which the iltum is conspicuously shorter than both skull
and femur, especially in the smallest specimen, NIGP
127586 (Currie & Chen, 2001), but it must be noted that
an ilium that is relatively small and shorter than the fe-
mur is found also in other compsognathids, including the
large and probably adult holotype of Sinocalliopteryx (Ji
et al., 2007a).

More precise data are available about the pubic foot.
The pubic foot of Scipionyx (Figs. 98, 102A), although
developed caudally, is comparatively less developed than
in other compsognathids considered ontogenetically older
(Chiappe, pers. comm., 2006). Size and shape of the pubic
foot change through ontogeny in Ceratosaurus (Britt et
al., 1999, 2000), in some oviraptorosaurs (Osmélska e?
al., 2004) and in Nedcolbertia, whose pubic foot is more
elongate craniocaudally in the subadult paratype (CEUM
5072) than in the juvenile holotype (CEUM 5071) (Kirk-
land et al., 1998). These data support the idea that ontog-
eny is responsible, at least in part, for the development of
the caudal process of the pubic foot. Based on Compso-
gnathus, ontogeny may have affected also the develop-
ment of the ischial foot, as it is more developed in the
French specimen, which, according to Peyer (2006), is
ontogenetically older than the German specimen.

Degree of ossification of the ilium - The ilium of
Scipionyx is well-defined in outline and well-ossified. Ac-
cording to Geist & Jones (1996), the extent of ossifica-
tion of the pelvis at hatching is a reliable indicator of the
precocial or altricial nature of a neonate archosaur. These
authors, in fact, show that the pelves of late foetal croco-
dilians and precocial birds are more ossified than those of
altricial birds, which, however, become active in the nest
in a matter of days after hatching, showing rapid postnatal
ossification of the pelvis. Based on these data, the only
information we can infer from the fully ossified ilium of
Scipionyx is that, if altricial, it was at least a few days old
at the time of its death.

118 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Non-ossified sternal plates - Contra Dal Sasso & Si-
gnore (1998a), Scipionyx lacks bony sternal plates (see
Humerus): it might have had, though, cartilaginous ster-
nal plates (see Ribs). Sternal plates are large and well-
ossified in adult Maniraptoriformes, but seem to be absent
in embryos of that clade (Norell ef a/., 2001). However,
there is no evidence of calcified sternal plates in theropods
outside the Maniraptoriformes, not even in adult individu-
als. Therefore, the condition exhibited by Scipionyx can
be considered ontogenetically uninformative.

Limb proportions - Compsognathid theropods have
forelimbs that are comparatively shorter than those in
several other coelurosaurs. In Compsognathus (Peyer,
2006), the forelimb measures 42% of the hindlimb. The
forelimb of Sinosauropteryx is shorter and stouter: ac-
cording to Currie & Chen (2001), it is less than 1/3 of
the length of the hindlimb (36% following Hwang e?
al. [2004]). The highest forelimb/hindlimb ratio among
compsognathids (i.e., 48%) is found in Huaxiagnathus
(Hwang et al., 2004). Gòhlich & Chiappe (2006) stated
that Juravenator has the longest forelimb among comp-
sognathids: precise measurements are not available but
it must be noted that, although the forearms seem to be
long, such a statement is based on the forelimb/hindlimb
comparison without taking into consideration that the
hindlimbs of Juravenator are rather short, possibly the
shortest among compsognathids. As mentioned, most
other coelurosaurs, including basal tyrannosauroids,
have comparatively longer forelimbs. For example, the
forelimb/hindlimb ratio is 60% in Guanlong (Xu et al.,
2006), 52% in Sinornithomimus (Kobayashi & Lii, 2003),
75-80% in almost all Laurasian dromaeosaurids (e.g., Xu
et al., 1999; Burnham et al., 2000) except 7ianvuraptor
(53%-Zheng et al., 2009), and even higher, i.e., 86-102%,
in the Aves (e.g., Elzanowski, 2002). Some other coe-
lurosaurs that are not strictly related to each other have
forelimbs as long as, or comparatively shorter than, those
of the compsognathids. Among these are tyrannosaurids,
alvarezsaurs including Ngwebasaurus (43%-de Klerk ef
al., 2000), the basal oviraptorosaur Caudipteryx (40%-
Zhou et al., 2000), troodontids (47%-Xu & Norell, 2004;
38%-Russell & Dong, 1993), some secondarily flightless
avians (Gauthier, 1986) and, probably, the Gondwanan
dromaeosaurids (e.g., Novas et al., 2008b).

In Scipionyx, measurements of the hindlimb are not
available because the specimen lacks the distal half of the
legs. With a forelimb/presacral ratio of 48%, it does not
differ considerably from Compsognathus (46%), although
the proportions are slightly different, Scipionyx having
comparatively longer skull and manus. The forelimb/
presacral ratio is only 39% of the length of the presacral
vertebral column in the NIGP 127586 specimen of Sino-
sauropteryx, roughly 10% less than in Scipionyx, empha-
sising the shortness of the forelimb of the former taxon.

Another, more consistent way to assess the proportions
of the Italian compsognathid by comparing forelimb and
hindlimb based on the preserved parts, is to look at the
femur/humerus ratio. This ratio is 1.42 in Scipionyx, about
1.75 in Sinocalliopteryx, 1.80 in the German Compso-

gnathus and in Huaxiagnathus (Hwang et al., 2004), about —

2.00 in Juravenator, 2.10 in the French Compsognathus
(Peyer, 2006) and 2.50 in Sinosauropteryx. In other coe-
lurosaurs, the ratio is 1.66 in Guanlong (Xu et al., 2006),

1.76 in Coelurus (Carpenter et al., 2005b), 1.80 in Tany-
colagreus (Carpenter et al., 2005a), 1.88 in Dilong (Xu et
al., 2004), 1.60 in Ornitholestes (Carpenter et al., 2005b),
1.52 in Sinornithomimus (Kobayashi & Lii, 2003), 1.69
in Sinornithoides (Russell & Dong, 1993), 1.93 in Mei
(Xu & Norell, 2004) and about 2.00 in Caudipteryx (Zhou
et al., 2000) and Nqwebasaurus (de Klerk et al., 2000);
on the other hand, it is 1.13 in Bambiraptor (Bunham ef
al., 2000), 1.10 in Sinornithosaurus (Xu et al., 1999) and
1.18 in Microraptor (Hwang et al., 2002). Thus, in Scipio-
nyx the humerus is comparatively longer with respect to
the femur than in any other non-avian coelurosaur except
the dromaeosaurids, which have by far the longest fore-
limbs among non-avian theropods. The ratio in Scipio-
nyx, however, is not much greater than that in some other
compsognathids such as Sinocalliopteryx; therefore, the
difference might be related to the ontogenetic stage of the
individual, and there possibly would have been positive
allometric growth of the hindlimb with respect to the fore-
limb. In Compsognathus, the difference between the Ger-
man and the French specimens is comparable to that be-
tween Scipionyx and Sinocalliopteryx. On the other hand,
both specimens of Sinosauropteryx described by Currie &
Chen (2001) show a similar ratio.

In conclusion, limb proportions and, in particular,
elongation of the humerus respect to the femur, distin-
guish Scipionyx samniticus from all other basal coelu-
rosaurs but are not useful in assessment of ontogenetic
stage, pending the discovery of other individuals of this
species representing different ontogenetic stages.

Carpal count and ossification - Scipionyx possesses
only 2, but well-ossified, carpal bones, which are tightly
and precisely articulated with the epipodial and the metapo-
dial elements, so much so that there is no space left for car-
tilaginous elements (Fig. 95). Based on this evidence, such
a definitive-looking arrangement would have been ac-
quired very precociously during ontogeny, or even during
embryogeny. Although early ossification of the carpals in
a bipedal predator may reflect a precocial functional util-
ity, the condition exhibited by Scipionyx is ontogenetically
uninformative. This issue is more extensively discussed in
the osteological description (see Carpus).

Elongation of the manus - The metacarpals of Scipio-
nyx are very long with respect to the epipodials (Fig. 94).
Although at first glance the long manus of Scipionyx ap-
pears as one of the most striking features of its skeleton
— raising the doubt that the proportions might be related
to ontogeny (Gishlick, pers. comm., 2000; Mateus, pers.
comm., 2004) — its manus is not comparatively larger than
in most other compsognathids (see Forelimb). Thus, the
ratio between the manus and the rest of the forelimb re-
sults ontogenetically uninformative, pending adult mate-
rial of this species.

Development of the fourth trochanter - The absence
of well-developed processes such as the fourth trochanter
may be indicative of the early ontogenetic stage of an in-
dividual. This correlation is clearly seen in the ornithis-
chian dinosaur Maiasaura, where bony attachments for
muscles form much later in response to muscle-induced
mechanical stresses on long bones (e.g., Geist & Jones,
1996). According to its phylogenetic position, however,

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

it is most parsimonious to hypothesise that the fourth tro-
chanter was absent or greatly reduced in Scipionyx, what-
ever the ontogenetic stage. The fourth trochanter is indeed
absent or greatly reduced in many small-bodied coeluro-
saurs, including compsognathids (see Femur).

Remarks - Summing up, most of the above-listed fea-
tures of Scipionyx are typical of immature animals, or, at
least, are representative of the condition expected to be
seen in an immature animal. Some of these features, such
as the presence of a frontoparietal fontanelle, no tooth re-
placement and no closure of the neurocentral sutures, sug-
gest that Scipionyx was a neonate, less than three weeks
old at the time of death.

This hypothesis calls for the re-interpretation of some
aspects of the palaeobiology and the soft anatomy of the
specimen. Concerning its palaeobiology, the presence of a
variety of prey (fish and lizards) in the gut of the neonate
might indicate that Scipionyx scavenged dead fish and liz-
ards right after hatching or, alternatively, might evidence
parental care, whether the animal was precocial or not. A
third hypothesis, that cannot be ruled out, is that Scipio-
nyx hunted actively for its prey, or at least some of it, but
it is difficult to imagine neonates such as the Pietraroja di-
nosaur being capable of catching fast-moving prey, such
as live fish and lizards, without parental assistance (see
also Palaeobiological Significance Of The Gut Contents
of Scipionyx). Interestingly, even in extant crocodiles,
which are without doubt precocial animals, parental feed-
ing has been observed a few times in captive specimens
(Huchzermeyer, pers. comm., 2010).

Regarding its soft anatomy, the unusual arrangement
of the intestine, with an empty space in the abdominal-
pelvic area (see Rectum), may actually be indicative of
the presence of a yolksac (Huchzermeyer, pers. comm.,
2010). In extant crocodilians, resorption of the yolksac
depends, to some degree, on food intake — the earlier the
hatchling starts feeding, the faster the resorption in the
3 weeks, circa, after hatching. In Crocodylus niloticus
hatchlings, the yolksac occupies practically the whole ab-
dominal cavity, becoming about half-size by the age of
2 weeks (Huchzermeyer, pers. comm., 2010). In extant
birds (chickens, geese and ducks), intensive absorption of

119

yolksac ingredients is observed during the first 5 days of
life, but residues of the yolksac can be found up to day
7 in more than 30% of chickens and 10% of geese; in
10% of chickens, yolksac remains can be seen even up to
day 16 (Jamroz et al., 2004). The empty abdominal-pelvic
space in Scipionyx samniticus is consistent with the space
occupied by the yolksac in a 2-week-old C. niloticus
that has died of yolksac retention (Fig. 110). In a normal
hatchling of that age, the yolksac would have been about
half that size. Considering a postulated faster growth rate
and, therefore, a faster metabolic rate for dinosaur hatch-
lings, one can infer the age of Scipionyx to be 3-7 days
old, maximum (Huchzermeyer, pers. comm., 2010). The
size of the supposed yolksac of Scipionyx is also compa-
rable to the maximum found in one-week-old extant birds
(Jamroz et al., 2004). This would not preclude that the
hatchling had had several meals.

Another interesting point is the overall body size of
Scipionyx: the Italian compsognathid has a skull length
of about 5 cm and an estimated body length of about 50
cm. Embryos from Lourinhà, referred to Lourinhanosau-
rus, are estimated to be 40 cm in length (Mateus e? al.,
1998); the maxilla of a post-hatching A//osaurus is 23 mm
in length (Rauhut & Fechner, 2005), suggesting that the
individual was 40-50 cm in length; the maxilla of an un-
hatched 7roodon embryo is only slightly smaller, 18 mm
long (Varricchio et a/., 2002); the skull of the oviraptorid
embryo is about 4 cm long (Norell et a/., 1994); the skull
of the Byronosaurus perinates is estimated to be approxi-
mately 5 cm long (Beaver & Norell, 2009); and the skull
of the therizinosaur embryo is less than 3 cm long (Kun-
drat ef al., 2008). Therefore, the holotype of Scipionyx is
within the size-range of the known late-stage embryos
and neonates of non-avian theropods. More interestingly,
considering that the holotype of Scipionyx was probably
slightly smaller at hatching, it may well have fitted into
eggs as large as those of 7roodon (12-16 cm x 5-6 cm
[Varricchio ef al., 2002]) or oviraptorids (18-19 cm x 6.5-
7.2 cm [Clark et al., 1999]). As a matter of fact, the 1:1
in ovo restoration of Scipionyx, based on its actual size,
i.e., disregarding that it was probably slightly smaller at
hatching, yields an 11x6 cm egg (Fig. 111). Among the
above mentioned theropods, it must be noted that the non-

Fig. 110 - A two-week-old Crocodylus niloticus died because of yolksac retention, before (A) and after (B) removal of the yolksac.
The position of the yolksac in the hatchling crocodile is consistent with the space left by the unusually displaced intestine of Scipionyx

samniticus. Scale bar = 40 mm. i : i
Fig. 110 - Un Crocodylus niloticus di due settimane morto per ritenzione del sacco del tuorlo, prima (A) e dopo (B) la rimozione dello

stesso. La sua posizione è compatibile con lo spazio lasciato dall’inusuale

metrica = 40 mm.

dislocazione dell’intestino in Scipionyx samniticus. Scala

| 20 CRISTIANO DAL SASSO & SIMONE MAGANUCO

coelurosaurian taxa Lourinhanosaurus and Allosaurus at-
tained an adult body size of more than 4 m and 7-9 m,
respectively (Mateus, 1998; Rauhut & Fechner, 2005),
whereas the adults of the remnant taxa (or the adults of
the most similar taxa in the case of the therizinosaurs),
which are coelurosaurs, were considerably smaller (7roo-
don, Byronosaurus and the oviraptorosaurs being 2 m or
less in total body length and therizinosauroids probably
related to the embryo being 2-3 m in length). Thus, tak-
ing into account 1) the size of the mentioned coelurosau-
rian eggs, embryos and respective adults, 2) the estimated
age of Scipionyx and its estimated size at hatching, and
3) the fact that compsognathids are among the smallest
non-avian theropods known and that Sinoca/liopteryx, the
largest known member of this taxon, measures 2.34 m in
total body length (Ji ef a/., 2007a), we may hypothesise
that adults of the Italian compsognathid did not exceed
this size (Fig. 112).

Although Sinocalliopteryx gigas is indeed the largest
of the compsognathids, being markedly larger than S7-
nosauropteryx or Compsognathus, it must be taken into
account that it is one of the few species probably repre-
sented by mature individuals. The fragmentary Aristosu-
chus pusillus is estimated to be about 2 m long (Holtz
et al., 2004) and Huaxiagnathus orientalis, represented
by an individual potentially approaching maturity, is esti-
mated to be 1.6 m long by Hwang et al. (2004) and about
1.9 m long by Peyer (2006). The probable compsognathid
Mirischia asymmetrica is also estimated to be about 1.9 m
long. Although immature, the French specimen of Comp-
sognathus, which is approximately as large as the largest
specimen of Sinosauropteryx, is estimated to be about 1.4
m long, and would have grown further before reaching
sexual maturity. Based on these observations, we think
that all compsognathids could have been able to lay eggs
matching the size of Scipionyx at hatching. Comparison
with extant non-flying birds suggests that in the ecologi-
cally equivalent non-avian theropods, a small adult body
size would not have precluded a relatively large egg size.

— E

Fig. 111 - Life-size in ovo restoration of Scipionyx samniticus, based on
its size at the time of death.

Fig. 111 - Ricostruzione in ovo di Scipionyx samniticus in grandezza
naturale, basata sulla taglia che aveva al momento della morte.

Fig. 112 - Body size of the hatchling Scipionyx samniticus (standing and in ovo) compared to the estimated body size of an adult indi-

vidual. Scale bar = 10 cm.

Fig. 112 - Dimensioni corporee (in piedi e în ovo) di Scipionyx samniticus neonato, confrontate con la taglia stimata di un individuo

adulto. Scala metrica =10 cm.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

PHYLOGENETIC ANALYSIS

The present phylogeny is based on the phylogenetic
data matrix of Senter (2007), which still represents one
of the most comprehensive phylogenetic analyses of the
theropod clade Coelurosauria published to date: 83 coe-
lurosaurian ingroup taxa vs 40 in one of the latest data
matrices (Choiniere e? a/., 2010, and references therein).
The analysis was performed in order to test the phyloge-
netic position of Scipionyx samniticus within the Coe-
lurosauria, and to see how the resulting topology would
differ from that of Senter (2007), after: 1) inclusion of
new taxa from recent studies and highly incomplete
taxa; 2) some changes concerning character coding in
the matrix; 3) inclusion/exclusion of characters possibly
related to ontogeny, in order to minimise the risk derived
by inaccurate identification of the ontogenetic stage in
fossil forms.

Modification of the data matrix

Terminal taxa - Besides Scipionyx samniticus, 9
more coelurosaur — or alleged coelurosaur — taxa were
added, thereby increasing the information content (95
taxa x 360 characters = 34,200 data points) of the Senter
(2007) matrix (85 taxa x 360 characters = 30,600 data
points). The new taxa in the matrix are Aniksosaurus
darwinii, Guanlong wucaii, Juravenator starki, Miri-
schia asymmetrica, Nedcolbertia justinhoffmani, Nqwe-
basaurus thwazi, Orkoraptor burkei, Santanaraptor
placidus and Sinocalliopteryx gigas. Although described
in preceding years, Mîrischia asymmetrica (Naish et al.,
2004), Nedcolbertia justinhoffmani (Kirkland et al.,
1998), Nqwebasaurus thwazi (de Klerk et al., 2000) and
Santanaraptor placidus (Kellner, 1999) were not in the
Senter (2007) matrix. Senter (2007) did not comment on
the exclusion of these taxa. Although all except Nqweba-
saurus are represented by highly incomplete specimens,
we decided to include them in the analysis. As recently
reported by Butler & Upchurch (2007; and references
therein), it has been demonstrated that even highly in-
complete taxa have the potential to provide significant
phylogenetic information, and taxa should only be ex-
cluded from a phylogenetic analysis a priori when their
deletion will have no effect on the inferred interrelation-
ships of the remaining taxa.

Concerning the other newly added taxa, Orkoraptor
burkei appeared in the literature in 2008, whereas the
remaining 4 taxa were not included by Senter (2007)
presumably because the papers, published around 2006-
2007 were in preparation at the same time as his own
paper.

The character states of Aniksosaurus darwinii were
scored according to Martinez & Novas (2006); those of
Guanlong wucaii according to Xu et al. (2006); those
of Juravenator starki according to Gòhlich & Chiappe
(2006), Gohlich et al. (2006) and unpublished photo-
graphs provided by Gòhlich & Chiappe in 2006; those
of Mirischia asymmetrica according to Naish ef al.
(2004) and photographs by Gohlich & Chiappe; those
of Nedcolbertia justinhoffmani according to Kirkland
et al. (1998); those of Ngqwebasaurus thwazi according
to de Klerk ef a/. (2000), and personal communications

and photographs by de Klerk (2004); those of Orkorap-
tor burkei according to Novas et al. (2008a); those of
Santanaraptor placidus according to Kellner (1999);
and those of Sinocalliopteryx gigas according to Ji et al.
(2007a). Following Norell et al. (2009), the taxon name
Saurornithoides junior has been replaced by Zanabazar
Junior, this taxon being distinet at generic level from
Saurornithoides mongoliensis.

Characters - Several changes were made to the Senter
(2007) matrix concerning wording, coding of character
states for some taxa, addition/deletion of some charac-
ter states and total replacement of characters considered
unclear. Few of these changes are due to disagreement
with Senter’s codings or to changes in the wording for
some characters, but most are due to: recognition of
new character states for other characters, availability
of illustrations and description of a new specimen of
Sinosauropteryx (Ji et al., 2007b), recent redescription
of Compsognathus (Peyer 2006), personal observation
on that specimen (MNHN CNJ 79) and personal com-
munications by Peyer (2005), as well as availability of
photographs of Coelurus and Huaxiagnathus (courtesy
of Gohlich and Chiappe). AIl these changes are listed in
Appendices 2-4 to make them easier to find for the sake
of evaluating or challenging them.

Other changes are due to the assessment of the ma-
turity-dependent characters (see Appendix 5), as some
of them proved not to be of taxonomic significance but,
rather, linked to maturity (i.e., to ontogenetic changes)
of the specimens. Whatever its precise ontogenetic stage
(see Ontogenetic Assessment), Scipionyx 1s not repre-
sented by mature individuals or individuals of relatively
late ontogenetic stage, as is the case in some other thero-
pod taxa. As pointed out by Tykoski (2005), treating
such taxa as if they were represented by adult specimens
could result in undeveloped (or under-developed) mor-
phologies being assessed for maturity-dependent charac-
ters that are expressed only in later stages of life. Coding
as unambiguously “absent” a character that is usually
expressed only late in ontogeny, ignoring that this is the
result of immaturity and not phylogeny, may eventually
alter the results of the cladistic phylogenetic analysis.
Similarly, it must be recalled that allometric trends and
heterochronic processes characterise the skeletal on-
togeny of all tetrapods, theropods included (Nicholls
& Russell, 1981; Long & McNamara, 1997; Rauhut &
Fechner, 2005), resulting in proportional differences be-
tween very young and fully mature individuals and hav-
ing a potentially negative impact on coding of characters
that compare an element’s size relative to another skel-
etal component (i.e., ratio-characters). Therefore, poten-
tially late-appearing, maturity-dependent characters and
ratio-characters must be carefully assessed for the taxon
(see Appendix 5) and, where necessary, scored as miss-
ing data in the absence of adult specimens, which may
or may not possess these characters or show definitive
skeletal proportions. Similarly, ratio-characters must be
carefully assessed also because of allometric differences
that are sometimes more related to the absolute size at-
tained by the taxon than to its phylogenetic affinities
(Nicholls & Russell, 1981).

109 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Multistate characters were considered as unordered,
unless they corresponded in all likelihood to a trans-
formation series. Multistate characters considered as
ordered are: ch17, ch18, ch38, ch41, ch46, ch66, ch77,
ch84, ch110, ch113, ch119, ch123, ch156, ch157, ch164,
ch169, ch172, ch175, ch176, ch200, ch217, ch277, ch298,
ch299, ch302, ch331, ch335 and ch345. A missing state
of a multistate character (due to the preservation of the
specimen) is coded as uncertain (e.g., 0/1). Multistate
characters have been coded as uncertain also when taxa
represented by immature specimens possess a character
state which could have been developed into a successive
state but not reversed into a preceding one during ontog-
eny (e.g., Appendix 5, ch169).

The two-stage approach recommended by Butler &
Upchurch (2007) was followed first. The application
of “safe taxonomic reduction” to a simplified version
of our data matrix (simplified to allow the program to
run), performed with the program TAXEQ3 (Wilkin-
son, 2001a) with the aim to identify taxa that can be de-
leted a priori without having an impact on the inferred
interrelationships of the remaining taxa, did not iden-
tify any taxa that could be safely removed. Redcon 3.0
(Wilkinson, 2001b) resulted as inapplicable because of
the high number of terminal taxa present in our analy-
sis. The parsimony analysis was therefore carried out
normally.

Results

A data matrix of 360 characters in 95 terminal taxa
(Appendix 6) compiled in NDE (Page, 2001) was ana-
lysed using the heuristic search of the most parsimonious
tree (MPT) of PAUP 4.0b10 (Swofford, 2002). Drawings
of the trees (Fig. 113) are based on the trees displayed
with Tree View (Page, 1996). The heuristic search was
performed using 1,000 random addition-sequence repli-
cates with ‘“maxtrees” set at 10,000. In total agreement
with the considerations of Kitching et al. (1998), no boot-
strap analysis was run.

The analysis generated 10,000 MPTs (L=1418 steps,
CI=0.3159, RI=0.7395 and RC=0.2336). The “tree de-
scription” option of PAUP was used to obtain the recon-
structed states for internal nodes, the character change
list and the apomorphy list (see Supplementary Informa-
tion).

Character transformation was optimised under both
accelerated transformation (ACCTRAN) and delayed
transformation (DELTRAN) options of PAUP. Character
state changes that shift from one node to another under
different optimisation regimes are listed as “ambiguous
apomorphies” while those that are tied to one node despite
differing the optimisation are listed as “unambiguous apo-
morphies” according to the terminology of Holtz (1994)
and Chiappe et al. (1996) followed by Yates & Warren
(2000) and Maganuco et a/. (2009). According to those
authors, the terms “ambiguous” and “unambiguous” are
not meant to imply the presence or absence of homoplasy
in the distribution of a given character state. This means
that an unambiguous synapomorphy may be reversed
deeper into the clade or may appear convergently in an-
other clade. Ambiguity may be caused by missing data or
incongruence in the data.

As this section is primarily concerned with the phylo-
genetic affinities of Scipionyx samniticus, we adopted the
same formal names and definitions for the clades within
the Coelurosauria as already adopted by Senter (2007).
The only exception is represented by the taxon Ornitho-
mimosauria. According to the phylogeny of Senter (2007),
Ornithomimosauria would be a junior synonym of Arcto-
metatarsalia. However, we prefer to refer to the most used
term (i.e., Ornithomimosauria), waiting for further studies
which will confirm this conclusion or not. To avoid prolif-
eration of taxon names, formal names and definitions are
therefore not provided for the groups not defined in those
papers such as Aniksosaurus + Nedcolbertia, the content
and the interrelationships of which have yet to be eluci-
dated and corroborated by further studies. The clade nota-
tion used below, “taxon X + taxon Y”, refers to the least
inclusive clade in the MPT2/majority-rule consensus tree
comprising the two given taxa, and does not imply that
these taxa share a direct sister-taxon relationship. Node
numbers are indicated in Fig. 113.

Comments - Despite several changes to the data ma-
trix (see above) and the addition of 10 taxa (Scipionyx
included), most of the results of this analysis agree with
those of Senter (2007), which in their turn agreed with
those of most other published numeric phylogenetic
analyses (Senter, 2007; and references therein). There-
fore, the nodes that in Senter (2007) show the same to-
pology are not discussed here, and a simplified version
of the cladogram is represented in Fig. 113. As men-
tioned above, the whole cladogram with node numbers,
the reconstructed states for internal nodes, the character
change list and the apomorphy list are available to the
reader as Supplementary Information. The present dis-
cussion refers to this simplified cladogram, emphasising
the position of the newly added taxa and the phyloge-
netic affinities of Scipionyx samniticus. The apomorphy
list is extensively reported only for nodes internal to the
Compsognathidae.

Mirischia asymmetrica, Aniksosaurus darwinii and
Nedcolbertia justinhoffmani are found to occupy a basal
position within Coelurosauria. Mirischia asymmetrica is
here the basal-most of the three species, whereas Anikso-
saurus darwinii and Nedcolbertia justinhoffmani form a
monophyletic group. Martinez & Novas (2006) consid-
ered Aniksosaurus a coelurosaur basal to Maniraptori-
formes (including tyrannosauroids) and more derived than
Ornitholestes and compsognathids, without testing its po-
sition in a phylogenetic analysis. We agree with Naish ef
al. (2004, and references therein), who found some comp-
sognathid affinities in Mirischia asymmetrica suggesting
that it was more closely related to Compsognathus than to
Sinosauropteryx. Their hypothesis, however, was not test-
ed in a phylogenetic analysis. According to the phylogeny
of Rauhut (2003), Mirischia (at that time “the Santana
compsognathid’”), Compsognathus and Sinosauropteryx
were indeed united into the clade Compsognathidae. This
result is not supported in the present phylogeny, being in
all likelihood influenced by the incompleteness of Miri-
schia, Aniksosaurus and Nedcolbertia, which lack parts
of the skeleton that might have supported their inclusion
in one of the more derived groups. As a matter of fact,
both Mirischia and Nedcolbertia were deleted a posterio-
ri by Butler & Upchurch (2007) because they resulted as

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

100
COELUROSAURIA

TYRANNOSAUROIDEA

COMPSOGNATHIDAE

99

99

99
MANIRAPTORA

OUTGROUP
Mirischia asymmetrica

n Aniksosaurus darwini
Nedcolbertia justinhoffmani
Eotyrannus lengi

88
Tyrannosaurus rex

% Gorgosaurus libratus

si Guanlong wucaii
92 Dilong paradoxus
Tanycolagreus topwilsoni
Coelurus fragilis
179 Juravenator starki
La) DA Sinosauropteryx prima
100 Huaxiagnathus orientalis
100 Compsognathus longipes
100 Sinocalliopteryx gigas
Li e Orkoraptor burkei
Scipionyx samniticus
ORNITHOMIMOSAURIA
Omitholestes hermanni
Santanaraptor placidus
THERIZINOSAUROIDEA
Nqwebasaurus thwazi
ALVAREZSAURIDAE
100 OVIRAPTOROSAURIA
100 TROODONTIDAE
100 È DROMAEOSAURIDAE
PARAVES AVIALAE

100 100

Fig. 113 - 50% majority-rule consensus tree of the most parsimonious trees (MPTs) generated by PauP 4.0.b10 (Swofford, 2002), based
on the data matrix given in Appendix 6. Percentages at nodes are indicated to the right of the nodes. The topology of this consensus
tree is identical to that of MPT number 2. Numbers in italic grey to the left of the nodes refer to the nodes of MPT number 2 that are
discussed in the text.

Fig. 113 - “50% majority-rule consensus tree” degli alberi filogenetici più parsimoniosi (MPTs) generati con PauP 4.0.b10 (Swofford,
2002) sulla base della matrice di dati pubblicata in Appendice 6. Le percentuali dei nodi sono indicate a destra dei nodi stessi. La
topologia di questo consensus tree è identica a quella del MPT n. 2. I numeri a sinistra dei nodi (in corsivo grigio) si riferiscono ai nodi

del MPT n. 2 discussi nel testo.

unstable taxa. In Holtz et a/. (2004), Nedcolbertia is allied
with a clade composed by Scipionyx and all the coeluro-
saurs more derived than the compsognathids. Guan/ong
wucaii is a tyrannosauroid found to be more advanced
than Di/ong paradoxus but basal to the Tyrannosauridae.
This is the only result which contrasts with the most re-
cent phylogenies (Xu et al., 2006; Li et al., 2009; Sereno
et al., 2009), where Guanlong wucaii is basal to Dilong
paradoxus, the former showing more plesiomorphic fea-
tures. Some of the features which acted as synapomor-
phies shared by Di/ong and the other tyrannosauroids
in those phylogenies are indeed not included in Senter
(2007). Their exclusion might be responsible for the more
advanced position of Guanlong wucaii here: this is an-

other clear indication that larger data sets with more taxa
are needed to test the phylogenetic hypotheses to the best
of our knowledge. Santanaraptor placidus is found to be
a basal maniraptoran that is more advanced than Orni-
tholestes. In the present phylogeny, Ngwebasaurus thwazi
is basal to the Alvarezsauridae. No phylogenetic analysis
was performed by de Klerk ef al. (2000), and Nqweba-
saurus was deleted a priori by Gòhlich & Chiappe (2006)
anda posteriori by Butler & Upchurch (2007), being con-
sidered an unstable taxon. In the phylogeny of Holtz et al.
(2004), Nqwebasaurus resulted to be a basal coelurosaur
more advanced than compsognathids, whereas Sereno
(2001) suggested that this taxon is part of a clade includ-
ing alvarezsaurids and ornithomimosaurs.

124 CRISTIANO DAL SASSO & SIMONE MAGANUCO

The affinities of the Compsognathidae and the apo-
morphies characterising both nodes internal to that taxon
and terminal taxa are listed below.

Node 183

Compsognathidae

Included taxa: Scipionyx + Juravenator

Unambiguous synapomorphies: 73 (0 + 2), external
mandibular fenestra absent; 109 (0 — 2), interspinal
ligament attachments in form of beak-like processes
below apex of neural spine; 206 (0 — 1), neural spines
on caudal dorsal vertebrae in lateral view craniocaudally
expanded distally, fan-shaped; 207 (0— 1), shaft diameter
of phalanx I-1 equal/greater than shaft diameter of radius;
299 (0 + 1), manual unguals II and II weakly curved.
Ambiguous synapomorphies under DELTRAN: 46 (1 >
0), dorsal surface of parietals flat, lateral ridge borders
supratemporal fenestra; 101 (0 + 1), cervical and cranial
trunk vertebrae strongly opistocoelous; 163 (0 > 1),
ridge bordering cuppedicus fossa terminates rostrally
to acetabulum or curves ventrally onto cranial end of
pubic peduncle; 185 (1 — 0), accessory trochanteric crest
distal to lesser trochanter absent; 237 (0 + 1), body of
premaxilla dorsoventrally shallow; 263 (0 + 1), cervical
prezygapophyses flexed; 284 (0 + 1), manual phalanx I-1
shorter than metacarpal II.

Ambiguous synapomorphies under ACCTRAN: 9 (0
— 2), basisphenoid recess between basisphenoid and
basioccipital absent; 32 (1 —> 0), postorbital process of
the jugal well-developed, taller than half orbit; 43 (0 +
1), rostral emargination of supratemporal fossa on frontal
strongly sinusoidal and reaching onto postorbital process;
53 (0 — 1), quadrate hollow, with depression on caudal
surface; 101 (0 + 1), cervical and cranial trunk vertebrae
strongly opistocoelous; 119 (0— 1), neural spines on distal
caudals absent; 120 (0 — 2), prezygapophyses of distal
caudal vertebrae strongly reduced as in Archaeopteryx
lithographica, 272 (0 > 1), wide distal expansion of
scapula present.

Remarks - The taxon Compsognathidae was erected
by Cope in 1871. It is defined as Compsognathus lon-
gipes and all taxa sharing a more recent common ances-
tor with it than with Passer domesticus. According to
Holtz (2000), it has proven difficult to find convincing
derived characters that unite compsognathids, setting
them apart from all other theropods, because the former
ones retain the basic, relatively unspecialised coleuro-
saurian body plan. For this reason, the Compsognathidae
have long been known only from Compsognathus. Sev-
eral new alleged compsognathids have been discovered
in recent years. The Compsognathidae were recognised
as the basalmost clade of coelurosaurs by Holtz et al.
(2004), including Compsognathus and Sinosauropteryx.
In the phylogeny of Rauhut (2003), the Compsognathi-
dae, including Compsognathus, Sinosauropteryx and
Mirischia, resulted as monophyletic and placed within
Coeluridae near the base of Coelurosauria. After a priori
deletion of 7 “highly incomplete” or “less relevant” taxa,
Scipionyx included, Gohlich & Chiappe (2006) found
a monophyletie Compsognathidae placed at the base of

Coelurosauria (tyrannosaurids excluded) and composed .

of Juravenator, Sinosauropteryx, Huaxiagnathus and
Compsognathus. The 4 taxa formed an unresolved poly-
tomy. Butler & Upchurch (2007) criticised the analysis

of Géhlich & Chiappe (2006), pointing out that the result
was strongly influenced by the a priori deletion. They
showed that in the total evidence strict component con-
sensus (i.e., without any deletion), the 4 mentioned taxa
plus Scipionyx resulted in an unresolved position at the
base of Maniraptora, in a more derived position than or-
nithomimosaurs. After a posteriori deletion of 3 unstable
taxa, they obtain a strict reduced consensus tree in which
Compsognathidae was a monophyletic taxon at the base
of Maniraptora and more advanced than Ornithomimo-
sauria, composed only of Compsognathus and Coelurus.
Huaxiagnathus and Sinosauropteryx resulted to be more
derived than Compsognathidae, whereas Scipionyx and
Juravenator resulted to be more basal.

Peyer (2006) considered Compsognathidae as con-
sisting of 7 taxa: Compsognathus longipes Wagner,
1861; Sinosauropteryx prima Ji & Ji, 1996; Huaxia-
gnathus orientalis Hwang et al., 2004; Mirischia asym-
metrica Naish et al., 2004; Aristosuchus pusillus Owen,
1876; Juravenator starki G6hlich & Chiappe, 2006;
Scipionyx samniticus Dal Sasso & Signore, 1998. Al
though her hypothesis needs to be supported by a pub-
lished phylogenetic analysis, Peyer (2006) carefully rec-
ognised numerous anatomical key features of the taxon
and proposed the following revised diagnosis: “absence
of an external mandibular fenestra, anteroposteriorly
expanded; dorsally fan-shaped mid-to-posterior dorsal
neural spines; hook-shaped ligament attachments on
dorsal neural spines; short, wide and only slightly in-
clined dorsal transverse processes; absence of pleuro-
coels in dorsal vertebrae; Mcl very stout, approximately
as broad as long; and proximal width of PhI-1 more
than minimal shaft diameter of radius”. In the present
analysis, 4 of these characters resulted in unambiguous
synapomorphies of the taxon (see above), and we agree
with Peyer (2006) that the remnant mix of characters is
also useful to diagnose the compsognathids and indicate
membership of this group, although some of these char-
acters are widely distributed within Theropoda. Among
these characters, as suggested by Currie & Chen (2001),
are: presence of a proportionally large skull; unserrat-
ed rostralmost but serrated lateral teeth; slender, hair-
like cervical ribs; and pubic foot with a limited cranial
extension (Martill et a/., 2000; Naish et al., 2001) but
caudally elongate, prominent ischial obturator process.
Other characters recalling the Compsognathidae and
a more generalised coelurosaurian bauplan (Ostrom,
1978; Gauthier, 1986; Holtz ef al, 2004) are: acute,
needle-shaped quadrate process of the quadratojugal;
stout, L-shaped lacrimal; large prefrontal; pronounced
scapular acromion; fan-like coracoid and caudally fac-
ing glenoid; manual claws with low curvature; and fee-
ble development of the fourth trochanter in the femur.

In the original analysis of Senter (2007), only 3 of
the taxa forming the Compsognathidae in the present
analysis were included: Sinosauropteryx, Huaxiagna-
thus and Compsognathus. They formed a monophyletic
Compsognathidae, more derived than Tyrannosauroi-
dea and less derived than Maniraptoriformes. Here, 4
of the newly added taxa, Scipionyx included, fall within
a monophyletic Compsognathidae (but see comments
about Orkoraptor, below), which results as composed
of 7 taxa. Finally, according to Naish et al. (2004),
Holtz et al. (2004) and Peyer (2006), the fragmentary

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

taxon Aristosuchus pusillus, although not included in
the present phylogeny or in the ones mentioned above,
is also supposed to be a compsognathid because of the
strongly similar form of the pubis, with its foot enlarged
caudally.

Node 178

Included taxa: Scipionyx + Orkoraptor

Unambiguous synapomorphies: 359 (0 + 1), cranial cau-
dal vertebrae with pneumatopores.

Ambiguous synapomorphies under DELTRAN: none.
Ambiguous synapomorphies under ACCTRAN: 21 (0
— 1), internarial bar flat; 49 (0 + 1), descending proc-
ess of squamosal does not contact quadratojugal; 52 (0
— 1), quadrate strongly inclined anteroventrally so that
distal end lies far forward of proximal end; 82 (0 — 1),
second premaxillary tooth markedly larger than third and
fourth premaxillary teeth; 85 (0 — 2), small number of
dentary teeth (<11); 93 (0 — 1), axial epipophyses large
and caudally directed, extend beyond postzygapophyses;
117 (0 + 1), box-like centra in caudals I-V; 132 (0 + 1),
hypocleideum on furcula present; 139 (0 — 1), humerus
longer than scapula; 184 (0 —+ 1), fourth trochanter on
femur absent; 231 (0 + 1), dentary teeth increase in size
rostrally; 244 (0 + 1), nasals shorter than frontals; 298 (0
— 1), manual ungual I weakly curved.

Remarks - Orkoraptor was described by Novas e?
al. (2008a) on the basis of a fragmentary skeleton from
the early Maastrichtian of Patagonia. According to their
majority-rule consensus tree, Orkoraptor resulted to be
a basal member of a monophyletic Compsognathidae.
Novas ef al. (2008a), however, pointed out that such re-
sult is unexpected and needs to be taken with caution, as
Orkoraptor is considerably larger than the characteristi-
cally small-sized compsognathids, it does not fit with the
Early Cretaceous biochron of the Compsognathidae, and
it lacks compsognathid synapomorphies other than the
unserrated mesial margin of each tooth and the lack of a
caudoventral process on the quadratojugal. Novas er al.
(2008a) also considered the rostrodorsally inclined rostral
process of the postorbital as a condition documented in
coelurosaurians more derived than tyrannosauroids. Ac-
cording to Benson et al. (2010), the postorbital of Orko-
raptor is almost identical to that of the allosauroid Aero-
steon, and in both taxa a pneumatopore is apparent on
the dorsolateral surface of the atlantal neural arch, and
the proximal caudal vertebrae are intensely pneumatised.
Based on those characters, Orkoraptor resulted to be
a neovenatorid allosauroid in the phylogeny of Benson
et al. (2010), which differs from the present phylogeny
and from the phylogeny of Novas ef a/. (2008a) in be-
ing focussed on non-coelurosaurian taxa. In the present
phylogeny, which is based on Senter (2007) and neither
includes non-coelurosaur taxa (with the exception of the
outgroups) nor some of the characters used by Benson ef
al. (2010), Orkoraptor resulted to be a basal member of
the Compsognathidae and the sister taxon of Scipionyx. It
must be noted, however, that none of the present comp-
sognathid synapomorphies and none of the synapomor-
phies shared with Scipionyx under ACCTRAN are coded
for Orkoraptor. Its position is therefore a consequence of
the presence of pneumatopores in the cranial caudal ver-
tebrae. Thus, despite our result, we do not refer Orkorap-
tor to the Compsognathidae because of the fragmentary

nature of the specimen, and the fact that this coelurosau-
rian phylogeny based on Senter (2007) is not adequate
either in characters or in number of non-coelurosaurian
taxa to test other possible affinities. Although the non-
coelurosaurian phylogenies (e.g., Benson et a/., 2010) do
not include an adequate number of coelurosaurian taxa
to definitely solve the question, we regard Orkoraptor
as a tetanuran theropod of possible neovenatorid affini-
ties. A certain degree of caution is needed, pending more
complete material and more comprehensive phylogenetic
analyses of Theropoda.

Orkoraptor

Unambiguous autapomorphies: none.

Ambiguous autapomorphies under DELTRAN: none.
Ambiguous autapomorphies under ACCTRAN: none.

Scipionyx

Unambiguous autapomorphies: none.

Ambiguous autapomorphies under DELTRAN: 8 (0 +
1), subotic recess (pneumatic fossa ventral to fenestra
ovalis) present; 21 (0 — 1), internarial bar flat; 43 (0 +
1), rostral emargination of supratemporal fossa on frontal
strongly sinusoidal and reaching onto postorbital process;
49 (0 + 1), descending process of squamosal does not
contact quadratojugal; 52 (0 —> 1), quadrate strongly
inclined anteroventrally so that distal end lies far forward
of proximal end; 82 (0 — 1), second premaxillary tooth
markedly larger than third and fourth premaxillary teeth;
117 (0 + 1), box-like centra in caudals I-V; 132 (0— 1),
hypocleideum on furcula present; 184 (0 > 1), fourth
trochanter on femur absent; 213 (0 — 1), quadrate cotyle
of squamosal open laterally exposing quadrate head; 231
(0 — 1), dentary teeth increase in size rostrally; 233 (0
— 1), height of skull (minus mandible) at middle of naris
less than half the height of skull at middle of orbit; 235 (0
— 1), cranial concavity of the preacetabular blade of the
ilium in lateral view present and slightly developed; 298
(0 + 1), manual ungual I weakly curved.

Ambiguous autapomorphies under ACCTRAN: none.

Remarks - On account of the large amount of anatomi-
cal details not available to previous authors, it is partly out
of place to compare the present phylogenetic position of
Scipionyx with those found in the few previous analyses
that included the Italian theropod (e.g., Holtz et a/., 2004;
Butler & Upchurch, 2007). Scipionyx was previously con-
sidered to be a basal member of the Maniraptoriformes
(Dal Sasso & Signore, 1998a; 1998b) of uncertain affinity
(Dal Sasso, 2001; 2003), then it was ascribed to the Coe-
lurosauria incertae sedis (Dal Sasso, 2004; Holtz e? al.,
2004). Peyer (2006) was the first to refer Scipionyx to the
Compsognathidae, based on the absence of an external
mandibular fenestra, the dorsal vertebrae dorsally fan-
shaped with a “hook-like” ligament attachment, a very
short Mel, and the proximal width of PhI-1 more than the
shaft diameter of the radius. Her assignment of Scipionyx
to the Compsognathidae, however, was not based on a
published phylogenetic analysis.

In the present phylogeny, Scipionyx samniticus
clearly results to be a compsognathid. Even though we
did not include in the matrix the characters we found
to be strictly ontogeny-related (Appendix 5), it remains
difficult to determine to what extent its basal position
within Compsognathidae is due to its generally plesio-

126 CRISTIANO DAL SASSO & SIMONE MAGANUCO

morphic condition in respect to the other compsogna-
thids, or to the early ontogenetic stage of the only known
individual.

Node 182

Included taxa: Sinocalliopteryx + Juravenator
Unambiguous synapomorphies: 28 (0 — 1), accessory
antorbital fenestra situated caudal to rostral border of
fossa; 41 (0 — 1), prefrontal greatly reduced in exposure;
286 (0 — 1), metacarpals II and III are appressed for their
entire lengths; 304 (0 + 1), shaft of ischium slenderer
than the pubic shaft.

Ambiguous synapomorphies under DELTRAN: 120 (0
2), prezygapophyses of distal caudal vertebrae strongly
reduced as in Archaeopteryx lithographica; 272 (0+ 1),
wide distal expansion of scapula present.

Ambiguous synapomorphies under ACCTRAN: 127 (1
— 0), lateral gastral segment shorter than medial one in
each arch; 157 (1 — 2), supraacetabular crest on ilium as
a separate process from antitrochanter absent; 213 (1 >
0), quadrate head covered by squamosal in lateral view;
257 (0 + 1), premaxillary teeth much smaller than the
maxillary teeth.

Sinocalliopteryx

Unambiguous autapomorphies: 39 (0 + 1), enlarged
foramen or foramina opening laterally at the angle of the
lacrimal, present; 142 (1 — 0), olecranon process weakly
developed; 166 (0 — 1), shaft of ischium, ventrodistal
end curved cranially; 243 (1 — 0), suborbital process of
Jugal short and dorsoventrally stout; 274 (1 — 0), humeral
length is half femoral length or less; 282 (1 > 0), medial
side of metacarpal II: expanded proximally; 283 (0 — 1),
metacarpal III<0.8x length of metacarpal II; 288 (1 >
0), length of manual phalanx II-2<1.2x length of phalanx
II-1; 346 (1 — 0), premaxillary teeth serrated; 347 (1 >
0), sublacrimal process of jugal dorsoventrally expanded
(taller than suborbital bar of jugal).

Ambiguous autapomorphies under DELTRAN: 9 (0
— 2), basisphenoid recess between basisphenoid and
basioccipital absent; 127 (1 — 0), lateral gastral segment
shorter than medial one in each arch; 235 (0 + 1), cranial
concavity of the preacetabular blade of the ilium in
lateral view present and slightly developed; 257 (0 +
1), premaxillary teeth much smaller than the maxillary
teeth.

Ambiguous autapomorphies under ACCTRAN: none.

Node 181

Included taxa: Compsognathus + Juravenator
Unambiguous synapomorphies: 154 (1 > 0), ventral edge
of cranial ala of ilium straight or gently curved; 349 (1 +
0), distal chevrons straight or L-shaped in lateral view.
Ambiguous synapomorphies under DELTRAN: 32 (0 +
1), postorbital process of the jugal reduced/absent; 93 (1
— 0), axial epipophyses absent or poorly developed, not
extending past caudal rim of postzygopophyses.
Ambiguous synapomorphies under ACCTRAN: 32 (0 +
1), postorbital process of the jugal reduced/absent; 171 (0
— 1), ischium 70% or less of pubis length; 235 (1 0),

cranial concavity of the preacetabular blade of the ilium -

in lateral view absent; 302 (1 + 0), manual phalanx III-
3 markedly shorter than combined lengths of phalanges
III-1 and III-2.

Compsognathus

Unambiguous autapomorphies: 33 (0 + 1), jugal qua-
dratojugal process beneath lower temporal fenestra rod-
like; 148 (1 — 0), distal carpals 1+2 well-developed,
covering all of proximal ends of metacarpals I and II;
169 (2 + 3), caudal process of the pubic foot present and
longer than 1/3 of the proximodistal length of the pubis;
175 (1 + 0), pubis propubic; 298 (0 — 1), manual ungual
I weakly curved; 348 (0 — 1), flexor tubercles of manual
unguals<1/3x height of articular facet.

Ambiguous autapomorphies under DELTRAN: 171 (0
— 1), ischium 70% or less of pubis length; 233 (0 + 1),
height of skull (minus mandible) at middle of naris less
than half the height of skull at middle of orbit.
Ambiguous autapomorphies under ACCTRAN: 127 (04
1), distal gastral segment longer than proximal segment.

Node 180

Included taxa: Huaxiagnathus + Juravenator
Unambiguous synapomorphies: 29 (1 — 0), tertiary
antorbital fenestra (promaxillary fenestra) absent; 71 (0
— 1), dentary with subparallel dorsal and ventral edges;
275(1-—0), length ofhumeral shaft between deltopectoral
crest and distal condyles< 4.5x shaft diameter.
Ambiguous synapomorphies under DELTRAN: 127 (1
— 0), lateral gastral segment shorter than medial one in
each arch; 157 (1 — 2), supraacetabular crest on ilium as
a separate process from antitrochanter, absent.
Ambiguous synapomorphies under ACCTRAN: 10!:(14
0), cervical and cranial trunk vertebrae amphiplatyan to
platycoelous; 233 (1-— 0), height of skull (minus mandible)
at middle of naris, more than half the height of skull at
middle of orbit; 240 (0 — 4), maxillary fenestra small and
round, not dorsally displaced; 327 (0 —> 1), diameter of
non-ungual phalanges of manual digit III<0.5xdiameter
of non-ungual phalanges of digit II.

Huaxiagnathus

Unambiguous autapomorphies: 272 (1 > 0), wide distal
expansion of scapula absent; 283 (0 — 1), metacarpal
IH<0.8xlength of metacarpal II.

Ambiguous autapomorphies under DELTRAN: 240 (0
— 4), maxillary fenestra small and round, not dorsally
displaced; 257 (0 — 1), premaxillary teeth much smaller
than the maxillary teeth; 327 (0 — 1), diameter of non-
ungual phalanges of manual digit III<0.5xdiameter of
non-ungual phalanges of digit II.

Ambiguous autapomorphies under ACCTRAN: 171 (1
0), ischium more than 70% of pubis length.

Node 179

Included taxa: Sinosauropteryx + Juravenator
Unambiguous synapomorphies: 267 (0 — 1), cranialmost
haemal arches <1.5* as long as associated centra; 274 (1
— 0), humeral length is half femoral length or less; 277 (1
— 0), length of radius<1/3x femoral length; 284 (1 > 0),
manual phalanx I-1 longer than metacarpal II.
Ambiguous synapomorphies under DELTRAN: 302 (14
0), manual phalanx III-3 markedly shorter than combined
lengths of phalanges III-1 and III-2.

Ambiguous synapomorphies under ACCTRAN: 257 (I
— 0), premaxillary teeth subequal in size to the maxillary
teeth; 286 (1 —> 0), metacarpals II and III are not appressed
for their entire lengths.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 1297

Sinosauropteryx

Unambiguous autapomorphies: 41 (1 + 0), prefrontal
large, dorsal exposure similar to that of lacrimal; 115
(0 —+ 1), caudal vertebrae without transition point; 118
(0 — 1), neural spines of caudal vertebrae separated
into cranial and caudal alae throughout much of caudal
sequence; 152 (0 — 1), digit I bearing large ungual and
unguals of other digits distinctly smaller; 226 (0 + 1),
distal humeral condyles on cranial surface; 345 (0 >
1), with fingers extended, tip of ungual I extends past
flexor tubercle of ungual II but does not extend past tip
of ungual II.

Ambiguous autapomorphies under DELTRAN: 53
(0 — 1), quadrate hollow, with depression on caudal
surface; 101 (1 — 0), cervical and cranial trunk vertebrae
amphiplatyan to platycoelous; 119 (0 + 1), neural spines
on distal caudals absent; 171 (0 + 1), ischium 70% or
less of pubic length; 286 (1 —> 0), metacarpals II and
III are not appressed for their entire lengths; 327 (0 >
1), diameter of non-ungual phalanges of manual digit
III<0.5xdiameter of non-ungual phalanges of digit II.
Ambiguous autapomorphies under ACCTRAN: 240 (4
— 0), maxillary fenestra large and round.

Remarks - According to Currie & Chen (2001), rela-
tively short forelimb, a prominent ulnar olecranon process
and a powerful manual phalanx I-1 are characters indicat-
ing a monophyletic Compsognathidae. These characters
likely represent specializations of Sinosauropteryx, the
taxon they described.

Juravenator

Unambiguous autapomorphies: 23 (1 + 0), caudal
margin of naris farther rostral than the rostral border
of the antorbital fossa; 39 (0 — 1), enlarged foramen or
foramina opening laterally at the angle of the lacrimal,
present; 84 (1 —> 0), maxillary and dentary teeth serrated;
85 (0 — 2), small number of dentary teeth; 120 (2 +
1), prezygapophyses of distal caudal vertebrae with
extremely long extensions of the prezygapophyses (up to
10 vertebral segments long in some taxa); 142 (1 + 0),
olecranon process weakly developed; 175 (1 — 0), pubis
propubic; 239 (1 — 0), rostral portion of the maxillary
antorbital fossa: small, from 10% to less than 40% of the
rostrocaudal length of the antorbital cavity; 298 (0 >
1); 321 (0 + 1), total length of pedal phalanx II-2 (not
counting posteroventral lip, if any)< 2x length of distal
condylar eminence; 346 (1 —> 0), premaxillary teeth
unserrated; 349 (0 + 1), distal chevrons straight upside-
down T-shaped.

Ambiguous autapomorphies under DELTRAN: 161 (1
0), brevis fossa shelf-like; 233 (0 —> 1), height of skull
(minus mandible) at middle of naris less than half the
height of skull at middle of orbit; 240 (0 + 4), maxillary
fenestra small and round, not dorsally displaced.
Ambiguous autapomorphies under ACCTRAN: 233 (0
— 1), height of skull (minus mandible) at middle of naris
less than half the height of skull at middle of orbit; 327 (1
— 0), diameter of non-ungual phalanges of manual digit
III>0.5x diameter of non-ungual phalanges of digit II.

SKELETAL TAPHONOMY

In this section we analyse the taphonomy of the speci-
men, focussing on the skeletal parts. We will dedicate
a specific chapter of the monograph (see Part II) to the
taphonomy of the soft tissues, together with some notes on
the diagenesis of local calcite areas of the embedding sedi-
ment that in our opinion were partly affected by the prox-
imity to the soft tissues, during the decay of the carcass.

As mentioned in the description, the position of fossili-
sation of Scipionyx cannot be taken as informative about
the cause of death. For one part, the lack of the distal por-
tions of the hindlimbs and of the tail does not depend either
on traumatic pre mortem events (e.g., predation) or on pos?
mortem scavenging, nor on taphonomical dispersal condi-
tions, because the rest of the skeleton is complete, articu-
lated and even associated to soft tissues of its own body.
On the other hand, the mouth of Scipionyx is likely open
because of bone dislocation and crushing during diagen-
esis, rather than to jaw-muscle contraction or to supposed
suffocation, as the bones forming the mandibular hinge are
apparently disarticulated and deformed. So, whether the an-
imal died in the water or was instead carried out to sea after
death, cannot be ascertained. As the head is upturned with
respect to the neck, but not in an “opisthotonic” condition
(sensu Ostrom, 1978), the carcass of Scipionyx could have
been exposed to sunlight for a short period at the most. Its
dehydration was likely caused by other factors, such as 0s-
mosis (see Soft Tissue Taphonomy), and probably occurred
after burial, when the animal was already trapped by the
weight of the sediments.

During the taphonomical processes, most of the cranial
bones of the left side, the left dorsal ribs and the left ele-

ments of the pectoral girdle slid in the same direction un-
der the counterlateral elements of the right side (Fig. 114).
This might indicate that the forces of diagenetic compres-
sion did not act completely vertically but had also an ob-
lique component. However, caudally to the caudal dorsal
vertebrae, all pelvic bones show a sliding direction that is
opposite to that of the paired presacral bones. It is very un-
likely that such a bi-directional dislocation, which occurred
in body regions distant from each other by only 1-10 cen-
timetres, was the result of two different diagenetic com-
pression forces that acted one opposite to the other. Thus,
we think that this sliding pattern reflects the lying down
(depositional) position of the specimen. The carcass might
have reached the sea floor and sunk in the substrate on one
side, but in an arched, inverted U-shaped passive pose, with
the mid-abdomen as the highest portion of the body, or vice
versa. This hypothesis is consistent with the position of
the centre of buoyancy of a bipedal vertebrate with a later-
ally compressed body, irrespective of whether the carcass
of Scipionyx sank to the seafloor by gravity after floating
or was dragged there by a current. Subsequently, sediment
compaction would have favoured the sliding of the paired,
relatively non-mobile bones in opposite directions.

It is interesting to note that in Scipionyx the compression
of the skull caused the palatine-pterygoid complex to move
towards the centre of the orbital and antorbital cavities, just
like in Juravenator (G6hlich & Chiappe, 2006: fig. 2a) and
Compsognathus (Peyer, 2006: fig. 4A). The rostral end of
the left dentary bent medially, forming a U-shaped turn that
is still in contact with the end of the right dentary (Fig. 42).
This indicates that the preserved medial bending of the den-

128 CRISTIANO DAL SASSO & SIMONE MAGANUCO

taries is natural, and that the two bones formed a U-shaped
symphysis. In fact, as we pointed out in the description (see
Dentary), it is unlikely that the symphysis of the young Sci-
pionyx was so firmly sutured to favour distortion of the right
dentary rather than separation of the two rami during diagen-
esis, taking also into account that Scipionyx is a hatchling
and the symphyseal contact is often relatively weak even in
sub-adult and adult theropods (e.g., Britt, 1991).

Regarding the ceratobranchials I, these bones are still
coupled albeit fossilised distant from the lower jaw. This
suggests that they moved when they were still united by
the hyoid cartilages.

The neck of Scipionyx is strongly bent backwards,
with the postzygapophyses slid onto the prezygapophyses
up to the maximum limit of craniocaudal sliding permit-
ted by their articular surfaces. The post mortem conditions
certainly favoured this bending and the related extreme
interlocking of the neural arches, but the neck is still artic-
ulated and in a position that could have been achieved in
life, and it is not comparable with the unnatural pos? mor-
tem bending visible in some fossil specimens (e.g., both
specimens of Compsognathus). The bending is present all
along the postaxial cervical series of Scipionyx, but the
neck appears more recurved at the base because the cra-
nial cervicals are fossilised in a lateral-laterodorsal view,
limiting the observer’s perception of the bending and giv-
ing a more rigid appearance to the neck with respect to,
for example, that of Sinocalliopteryx.

In life, the distal extremities of the cervical ribs were
probably not so tightly appressed to the centra. However,

soft tissue dehydration occurring after death caused the
ribs to adhere to the vertebrae, and the favourable neck
posture and the structural flexibility of the ribs limited
their fracture.

As reported in the description, the neural arches are
well-articulated all along the vertebral column, with the
zygapophyses interlocking: their conformation has kept
them linked together. However, most of the centra (i.e.,
C5, C7-10, DI, D3-11, S4-S5 and Cal-5) — which are not
yet fused to their arches, not held by the apophyses and
definitely more inclined to move and roll because of their
shape — have detached from the arches in various ways. It
is interesting to note that, in the caudal series, the centra
have detached from the arch in a manner very similar to
the holotype of Huaxiagnathus (Hwang et al., 2004: fig.
4C). The D7 neural arch of Scipionyx shows a strange
plastic deformation: instead of it being straight, the dorsal
margin of this bone shows a S-shaped bending at the level
of the diapophysis, despite it being reinforced by a ridge.

In both manus, the position of fossilisation of the first
digits compared to the others gives an interesting insight
into the functional anatomy of the Italian compsognathid
(see Pectoral Girdle And Forelimb in Skeletal Recon-
struction And...).

Summing up, the cause of death of Scipionyx cannot
be hypothesised with this set of data. The mode of pres-
ervation of the skeleton indicates that the carcass, rather
than undergoing violent crushing, was subjected to slow,
plastic diagenetic deformation after rapid and complete
burial in a soft substrate.

Fig. 114 - Interpretive drawing of the passive pose of the carcass of Scipionyx samniticus on reaching the sea floor (A), based on the
relative position of the left (red) and right (green) paired bones in the fossil skeleton as seen from below (B).

Fig. 114 - Ricostruzione ipotetica della posa passiva della carcassa di Scipionyx samniticus al momento della deposizione sul fondale
marino (A), basata sulla posizione relativa, nel fossile visto dal basso (B), delle ossa pari del lato sinistro (rosse) e destro (verdi).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 129

PART II - SOFT TISSUE ANATOMY

Introduction

The notoriety, even at the popular level, of Scipionyx
is primarily due to the preservation of soft tissues and an-
atomical detail, including exquisitely fossilised internal
organs, never seen before in any other dinosaur. Its level
of preservation was already considered exceptional at the
macroscopic level, so much so that Scipionyx represents
an essential reference point for reconstructing the inter-
nal anatomy of dinosaurs as a whole (e.g., Leahy, 2000;
Paul, 2002; Chinsamy & Hillenius, 2004).

The employment of more sophisticated investiga-
tive methods, including examination under UV light and
SEM, has revealed that the soft tissues of this 110-mil-
lion-year-old fossil are not simply imprints but mineral-
ised in three dimensions and are exceptionally preserved
even at a cellular, and, in some instances, a subcellular,
level. Moreover, SEM element microanalysis has re-
vealed the presence of remains preserved as thin organic
films of endogenous chemical compounds derived from
the decay of the dinosaur’s carcass. Here, we describe
the astounding level of preservation not only of previ-
ously seen organic remains but also of those that have
been newly detected, most of which are unique to the

fossil record. However, despite the high degree of pres-
ervation of Scipionyx there are no traces of scales, feath-
ers or other integumentary structures. This unique kind
of fossilisation, whereby internal organs are preserved
but the integument is not, depends upon the particular
physico-chemical conditions of the aquatic environ-
ment in which the dinosaur became fossilised; this will
be examined later. Evidence of the high preservational
potential of the Pietraroja outcrop dates back two centu-
ries to when the fossil locality was first described. Costa
(1853-1864), for example, reported the exceptional fos-
silisation of skin and cartilagineous parts of a guitarfish
of the genus Rhinobatus (Fig. 14D). Among reptiles,
at least two specimens of lepidosaurs were found with
preserved remnants of soft tissue: the holotype of the
squamate Chometokadmon fitzingeri is pictured by Co-
sta (1853-1864) as broadly covered with patches of skin,
the squamation of which is clearly visible on the skull
and neck (Fig. 16); an indeterminate rhynchocephalan
recently described by Evans ef a/. (2004) preserves an
intestinal tract, linking the remnants of an ingested liz-
ard to a faecal mass ready to be voided. So, without any
doubt, the Pietraroja Plattenkalk represents a Konservatt-
Lagerstàtte (sensu Seilacher, 1970).

INTERNAL SOFT TISSUES

As extensively documented herein, all the internal soft
parts of Scipionyx, except for the oesophagus and the pur-
ported liver, are three-dimensionally mineralised tissues,
not simply imprints. The most important internal soft tis-
sues, which had not yet been examined in detail at the
time, were first described by Dal Sasso & Signore (1998a,
1998b). These include skeletal muscle remains at the base
of the neck and in the proximal region of the tail, connec-
tive tissue associated to the muscle bundles in the same ar-
eas, tracheal rings close to the furcula, faint traces of what
might have been the liver, and the entire intestine — despite
some claims, no heart is preserved in Scipionyx, at least not
with its original morphology (see Liver And Other Blood-
Rich Organs). These relatively large structures emerge
in relief and are clearly visible to the naked eye; they are
perfectly distinguishable from the skeletal elements, also
on account of their distinctive ochre colour, intermediate
between the brown bones and the grey-yellowish matrix
(Fig. 115). Other organic remains are preserved as thin
films, which can be seen under UV light as fluorescent ar-
eas of different colours (Fig. 116). With the exception of a
peculiar dark blue-purple colour marking the hepatic and
other blood-rich remnants, the soft tissues exhibit a bril-
liant golden-yellow hue. In addition to the soft tissues for-
merly described by Dal Sasso & Signore (1998a, 1998b)
and later by Dal Sasso (2003, 2004), traces of oesophagus,
blood vessels, visceral muscles, other small patches of
skeletal muscles, cartilages and ligaments have been sub-
sequently found and are discussed herein.

Besides Scipionyx, three-dimensional preservation of
muscle remains and other non-osseous internal soft tissue
has been described only in six, possibly seven, other dino-
saur specimens. First were two Brazilian theropods of simi-

lar age (Albian, Romualdo Member of Santana Fm.), fos-
silised in calcareous concretions (Kellner, 1996a, 1996b).
One of the specimens, later named Mirischia (Naish et al.,
2004), preserves a short tract of intestine in association with
the pelvic bones (Martill et a/., 2000); the other, later named
Santanaraptor (Kellner, 1999), preserves remains of skin
and muscle fibres. In describing the latter, Kellner (1996b)
stated that “since epidermis and muscle fibres are preserved
in three dimensions, this might be the best fossilised soft
tissue of a dinosaur known so far”. However, the compari-
son of analogous soft-tissue structures shows that Scipionyx
is better preserved: for instance, Scipionyx°s blood vessels,
which are either encased within bones or embedded in
phosphatised soft tissue, are not simply channels or rod-like
structures that have been secondarily filled in — they have
retained their walls as a distinct texture (see below).
Mineralised dinosaurian soft tissue was also unearthed
in the Lower Cretaceous deposits of Las Hoyas, Spain: a
single specimen of the ornithomimosaur Pelecanimimus
has fossilised in a lacustrine lithographic limestone, and has
preserved skin and muscle with some three-dimensional
detail, replicated in an iron carbonate (Briggs et al., 1997).
In 2000, the discovery of a fossil dinosaurian heart
was announced (Fisher e? a/., 2000). The putative heart,
including purported remnants of the aorta, was preserved
as a nodule of iron minerals inside the thorax of an orni-
thischian of the genus Thesce/osaurus (nicknamed “Wil-
lo”) from the Upper Cretaceous of South Dakota, USA.
The authors suggested that the organ had been saponified
under anaerobic burial conditions, and then changed into
goethite. Computed tomography imagery showed a four-
chambered structure, corresponding to the two atria and
the two ventricles of the heart of mammals and birds.

CRISTIANO DAL SASSO & SIMONE MAGANUCO

130

131

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

UNBJEY ke)

SIALLVAIMIC TVNYACIdI

S|jessan poojq E

ANSSIL UWINOSNW + FAILDANNOI
SAJOSNUI [EJBOSIA Si

sejosnw opewos pr

SAINSSIL YVTINIOSNW
anssi) YOL-poojq
/ anaiodojewsey
ves
abennueo snoJqy
sSenueo suyeAy
Wna]}soned
squewebi| pue suopus]

#

(DI
seuejuasaw pue ennuenpe iI

|

(snoJqy Jo asoo]) anssi
BAIOGUUOI 2JEU]ULJE)APuI

@NSSI) AANIGUUOI AS00]

SINSSIL FALLI INNO

SIIHODILVI
1V9I90101SIH

— 4 (89L-S91)

; anssK yos
7 smejo AuJoOy

0) pa}ejai Jou a]iojeo

SUIEWUISI DIUEBIO 0) pa]ejai
Ajgissod sJa)snjo apojeo

sjua}uo9 nb E

(p8L ‘6EL-2£1)
Ueaug ‘uaajdsi, ‘Jani|

(pSL ‘ESL) ajosnw
|euowajolyosiognd

(971 ‘2z1) sabennueo
Jejnoie Jejnoipuadde

(ESL ‘OGL-8pL
‘Gp|) Wnjoal

(SEL ‘PELI ‘6z1)
B2UY9eI}

(LLL ‘SEL ‘6Z1)
snbeydosao

a

LTS

(6zL) snwAW]L
‘@NSSI) 2AI]OSUUOIÌ,

(pzL) fa
squawebil jenxce (ESI ‘OGL ‘SpI)

]ajjad |ed9E,

1,

D
TE
|
\)

(LOL 'GEGL-SSL'ESI)
sajosnui |estoLUaJOpneo

:

)
sajosnwu yoau

(791) uawebili
‘anssi) 2AI]9BUU0I
|eixedAy |jepneo

(SyL ‘EHL-OpL)

(9£1) sajosnw unuaponp

|eIxeda |esJop

(ZL ‘LS1) SISSA
Poo]q 9119}UasaLU

- ununfal
al naar eva (L9-Sp1) fol

(971) sabemueo Jejnoie |elxe
(CAL'L9L-8SL ‘£SL) eidas ajosnwi

IEPNEIOIYISI-OII

(O9L-8SL ‘ESILI ‘SEL ‘6ZL)
aNSSI} AAIOSUUOII, S}euILUI9}apuI

(ezIL ‘Zz1) suew1 |ea}solJad

|

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY | 33

Fisher et al. (2000) interpreted this structure as indicating
an elevated metabolic rate for Thescelosaurus. However,
modern crocodilians and birds both have four-chambered
hearts, so dinosaurs probably had one as well, but the dif-
ferent metabolism of the two living clades of archosaurs
indicates that the heart structure is not necessarily tied
to metabolic rate (Chinsamy & Hillenius, 2004). What's
more, the interpretation of the ferrous nodule as a fossil
heart is not universally accepted: according to Rowe er al.
(2001), the heart is actually a concretion. As they note, the
anatomy given for the object is incorrect (for example, the
“aorta” narrows coming into the “heart” and lacks arteries
coming from it), and the concretion partially engulfs one
of the ribs and has an internal structure of concentric lay-
ers in some places; another concretion is preserved behind
the right leg.

Murphy et al. (2007) reported three-dimensional mus-
cle remains in the ventral portion of the neck, in the right
shoulder and in the tail of a hadrosaurid of the genus
Brachylophosaurus, coming from the Upper Cretaceous
of Montana, USA. The specimen, nicknamed “Leonardo”,
preserves integumental remains over 90% of its body, as
well as some gastric contents.

Manning (2008) described preliminarily the finding
of another partially intact hadrosaurid mummy, provi-
sionally nicknamed “Dakota”, unearthed in 1999 in the
Upper Cretaceous of North Dakota, USA. The fossilised
remains, which include skin, muscle, tendons and liga-
ments, represent the first-ever find of a dinosaur where
the skin “envelope” had not collapsed onto the skeleton,
allowing to calculate muscle volume and mass.

A seventh possible finding of dinosaurian soft tissue
was reported by Chin er al. (2003), who identified undi-
gested muscle and connective tissue of a possible pach-
ycephalosaurid dinosaur within a tyrannosaurid coprolite
from the Late Cretaceous of Alberta, Canada.

Lastly, Carrano & Sampson (2008) mentioned the
presence of muscle impressions and other soft tissue re-
mains in the abelisaurid theropod Aucasaurus from the
Upper Cretaceous of Argentina.

In any case, none of the specimens above provides
as much information on the anatomy of internal organs
as does Scipionyx samniticus: even before the present
monograph, the little Italian theropod has been a refer-
ence specimen for the reconstruction of the position of
the thoracic and abdominal organs, for supporters (e.g.,
Leahy, 2000; Paul, 2001, 2002) and opposers (e.g., Ru-
ben et al., 1999, 2003; Chinsamy & Hillenius, 2004) of
the avian origin of dinosaurs, alike. We relegate to a spe-
cific chapter our opinion on this matter, as we intend first
to describe, with the utmost accuracy, what in actual fact
can be seen in the specimen. Detailed examination of the
holotype of Scipionyx samniticus allows to identify a va-
riety of soft tissues, some of which are still associated
into whole, or parts of, organs. For this reason, and for
ease of description, we have subdivided this chapter into
sections dedicated to the best preserved anatomical parts,
topographically ordered in a craniocaudal direction, and
beginning with the innermost tissues, i.e., the ones most
intimately related to the skeletal elements. The histologi-
cal attribution to individual tissue categories and/or to a
pool of tissues composing each organ has been dealt with
in the discussion part of each chapter and in the synoptic
table of the soft tissues (Fig. 117).

Soft tissue within the bones

The presence of soft tissue remains within the bones
was investigated through SEM analysis of a microsample
taken from the dorsal margin of the shaft of the right third
dorsal rib. The sample contains blood vessels and shows
a cellular morphology, with a level of preservation that,
presumably, is present in the whole skeleton.

The most informative image shows a portion of the
compact bone layer of the rib in oblique section with a
blood vessel going through it (Fig. 118A). The lamellae
are 1.5-2.0 um thick and show a fine-grained texture; here
and there, the interlamellar lines accommodate lenticular
spaces which, by comparative anatomy (e.g., Kardong,
1997), represent the lacunae of the osteocytes. On closer
examination (Fig. 118B), the lacunae preserve subcellular
details such as a finely zig-zag shaped margin, made by al-
ternating bony papillae and thin intrusions into the lamel-
lae, which represent the canaliculi, i.e., the ultrastructural
lacunae left by the net of the filipodia. The fine intercon-
nections among the densely branched canaliculi is likely
responsible for the fine-grained texture of the lamellae.
This is confirmed by comparison with published pictures
of dinosaur osteocytes (Schweitzer ef a/., 2008: fig. 2),
in which lenticular cells are interconnected by peculiar
branched filipodia.

Pointed bony structures, protruding from the cutting
plane but well-rooted in the compact bone by branched
pillars, may represent transitional connections with the
spongy part of the rib or, simply, strengthening structures
(Fig. 118A-B). The blood vessel cut by the sampling has
an oval cross-section, measuring 3.0-5.5 um in diam-
eter, and preserves its wall in the form of a constantly
thick (0.5 um), differently textured, lighter coating (Fig.
118C). The lumen, virtually undeformed by diagenesis,
appears paved by a smooth, wavy layer (possibly, the en-
dothelial cells).

Soft tissue and cellular remains within cortical and
medullary dinosaur bone were recently investigated by
Schweitzer et al. (2007, 2008), who, after proper dem-
ineralisation, found remarkable preservation of flexible
and fibrous bone matrix, hollow blood vessels, intravas-
cular material and osteocytes: this demonstrated that, ex-
ceptionally, a fossil bone is capable of encasing and pre-
serving soft tissue for millions of years.

With respect to these unique results, we will comment
at least on the most comparable elements, i.e., the blood
vessels. In Scipionyx, although these structures have been
preserved in the form of a mineralised tissue, they are
present not only within the undoubtedly protective bone
but also in the soft tissue of the intestines (see below).
In both locations, the vessels are hollow, even though no
demineralisation or any other chemical cleaning technique
has been applied. Anticipating criticism on the part of the
reader, we would like to point out that the structures we
regard as being the walls of blood vessels do not represent
a mineral precipitate or a human artefact. In the first case,
one would expect nearby hollow structures to be similarly
lined with precipitate deriving from interstitial mineral
water; however, it is evident from the SEM images that
this is not the case, at least in this portion of the examined
bone (Fig. 119). With regard to a human artefact, two facts
rule out any chance of recent organic contamination: first,
we know that the person who discovered and first prepared

134 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 118 - Soft tissue preservation at the cellular level within the bones of Scipionyx samniticus. A) SEM image of a dorsal rib, show-
ing a portion of the compact bone layer in transverse section. Besides the lacunae of the osteocytes (B), a blood vessel is exposed and
shows preservation of its wall (C). See Appendix 1 or cover flaps for abbreviations.

Fig. 118 - Conservazione dei tessuti molli a livello cellulare, all’interno delle ossa di Scipionyx samniticus. A) immagine SEM di una
costola dorsale, che mostra una porzione di tessuto osseo compatto in sezione trasversa. Oltre alle lacune degli osteociti (B), è esposto
un vaso sanguigno con la sua parete ben conservata (C). Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

the fossil used an industrial chemical product (polyester
resin), but only for assembling three fractures of the origi-
nal slab — one crossing the skull, one crossing the pelvis,
one crossing the tibiae (G. Todesco, pers. comm., 2003)
— and applied as a paste to the fossil’s surface, so without
a permeating capability (Fig. 9); second, after Todesco’s
work, the specimen was restored and further prepared by
a collaborator of the Museo di Storia Naturale di Milano
(S. Rampinelli) along with one of the authors of this study

Fig. 119 - SEM image documenting that, in the vicinity of the blood
vessel shown in Fig. 118A (shaded area), even the largest bone vacui-
ties (arrows) are not lined with any mineral precipitate or recent organic
contaminant.

Fig. 119 - Questa immagine SEM documenta che, vicino al vaso san-
guigno mostrato in Fig. 118A (area ombreggiata), anche le cavità ossee
più grandi (frecce) non sono tappezzate da alcun precipitato minerale,
né da inquinanti organici recenti.

(CDS), who lightly impregnated the fossil with 5% Par-
aloid-B72 in acetone to protect the exposed surface of the
fossil. For this reason, the samples of Scipionyx that we
describe here were analysed along with a set of control
samples taken from fragments of a fossil fish found in a
similar layer of the upper Plattenkalk at Pietraroja, that had
been previously impregnated with the same preservative
(5% Paraloid-B72). SEM element microanalysis revealed
immediately the presence of the preservative on the por-
tions of control samples impregnated with Paraloid-B27,
as a medium-high carbon (C) peak (Fig. 120). This peak
was absent in all the microanalyses we conducted on the
samples taken from the rib of Scipionyx, even when we
aimed the microprobe at the cross-sectioned wall of the
blood vessel (Fig. 121A) or the surrounding bone (Fig.
121B). The C peak is also absent or low in most SEM-ana-
lysed soft tissues of Scipionyx (see below), indicating that
the preservative we used did not contaminate the specimen
in depth, but it is limited to the exposed surfaces. This con-
firmed the validity of our choice of always analysing the
reverse side of the samples.

Summing up, our SEM element microanalysis revealed
that the blood vessel in the rib of Scipionyx is composed
mostly of calcium phosphates, with proportions strictly
similar to all other phosphatised tissues, which clearly be-
long to the dinosaur and, thus, cannot be alien to its body
(see peaks of all soft tissues, in the following chapters).
The same elemental ratios occurred when analysing other
blood vessels preserved in the specimen (see Intestine).
Therefore, these vessels, their hollow lumina and their
finely ridged walls, which are of uniform thickness but
tapering at branching points, cannot be artefacts derived
from contamination with preservatives, such as polyvinyl
acetate, nor any other organic, equally carbon-rich exog-
enous intrusion, such as fungal invasions or extracellular
polymeric substances secreted by microbes.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 135

Ca

or=13.965 ke 1 ent ID=Nphl
1000 Window 0.005 - 40955= 27658 ent A

sor=15 525 keV Ocnt ID=
B Vert=2000 Window 0.005 - 40.955= 23511 ent G

1664Mm

Fig. 120 - SEM element microanalysis of a control sample (bone of a fish from the Pietraroja Plattenkalk) impregnated with the same
preservative (5% Paraloid B72) used on Scipionyx samniticus, and then cross-sectioned. Note the carbon peak (C) and the amorphous
cuticle (B, arrows) produced by the preservative, which are both much lower/absent in the internal portions of the sample (A).

Fig. 120 - Microanalisi degli elementi al SEM di un campione di controllo (tessuto osseo di un pesce del Plattenkalk di Pietraroja)
impregnato con lo stesso consolidante (Paraloid B72 al 5%) usato su Scipionyx samniticus, € poi sezionato. Si noti il picco del carbonio
(C) e la cuticola amorfa (B, frecce) originati dal consolidante, che sono molto inferiori/assenti nelle porzioni interne del campione (A).

(©)
do

T

È
=

—
{=}

‘ursor=35.405 keY Ocnt ID=
Vert=1000 Window 0.005 - 40.955= 29811 cnt

dA

FP]

i rusor=14.685 keV Sent ID=

o Vert=1000 Window 0.005 - 40.955= 40690 cnt

6% E G
n

#
»
È

ae

Fig. 121 - SEM element microanalysis of the wall of a blood vessel (A) encased in the sampled rib of Scipionyx samniticus (B), com-
pared with element microanalysis of the surrounding bone tissue (C). Both are composed of calcium phosphate and are devoid of any
recent carbon-rich, natural or synthetic contaminant.

Fig. 121 - Microanalisi degli elementi al SEM della parete del vaso sanguigno (A) incorporato nella costola campionata di Scipionyx
samniticus (B), a confronto con la microanalisi del tessuto osseo circostante (C). Entrambi sono composti da fosfato di calcio e sono
privi di qualsiasi contaminante organico, sia naturale che sintetico.

Periosteal remains

Some bone surfaces are covered with a yellowish
patina that gives them an opaque aspect when observed
with the naked eye or under the light microscope. As
experimented during the mechanical preparation of the
specimen (Dal Sasso, pers. obs, 1994-1997), this patina
is almost as hard as the underlying bone, so is not consist-
ent with the kind and the amount of preservative we ap-
plied (see above). More importantly, we are sure that the
patina was present prior to any human activity, because
we found it even on the portions — such as the cranial
sides of the ischia — brought to light subsequently by
one of us (Dal Sasso, pers. obs., 2004). Indeed, the patina
is more extensively preserved in these newly exposed ar-

eas, suggesting that it was likely widespread in the origi-
nal fossil. Under Wood”s lamp, the areas covered with
this patina gain an opaque brown colour that is lighter
than the dark brown of adjacent bone surfaces.

In our opinion, this patina represents residual patches
ofthe periosteum, a thin sheath of dense fibrous connective
tissue that, in vivo, coats the cortical layer of all bones. As
shown in the synoptic table of the soft tissues preserved in
Scipionyx (Fig. 117) and from some exemplifying photos
(Figs. 122-123), the most extended periosteal patches are
preserved on the right nasal, right frontal and lefi splenial
of the skull, and on the right scapular blade, right pubic
shaft, left pubic apron, right proximal and distal femoral
shafts, right fibula and, more extensively, on both ischial
shafts, in the appendicular skeleton.

136 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 122 - Close-up of the right dentary of Scipionyx samniticus, show-
ing periosteal remains in the form of a yellowish opaque patina. Scale
bar= 1 mm.

Fig. 122 - Particolare del dentale destro di Scipionyx samniticus, con
residui di periosteo sotto forma di una opaca patina giallastra. Scala
metrica = 1 mm.

Axial ligaments

Thin strips of an ochre-yellow material having a con-
sistency intermediate between the matrix and the bone can
be found in few, quite limited, points along the vertebral
column. Although no particular microstructure can be dis-
tinguished under the light microscope, the arrangement of
these strips, which are located on the apophyses of some
neural arches, suggests that they may represent remnants
of intervertebral connecting structures. The most cranial
structure is located on the lateral wall of the prezygapo-
physis of the 9! cervical vertebra, close to the prezyga-
pophyseal articular surface (Fig. 124). This position cor-
responds to the zygapophyseal articular capsule, which
is present, Just in the formerly described point, in extant

Fig. 123 - Patches of periosteum are well-preserved on the right pubic
bone of Scipionyx samniticus (A), in particular on its shaft (B). Scale
bar = 0.5 mm.

Fig. 123 - Lembi di periosteo sono ben conservati sull’osso pubico
destro di Scipionyx samniticus (A), in particolare sulla sua diafisi (B).
Scala metrica = 0,5 mm.

crocodilians (Frey, 1988) and birds (Baumel & Raikov,
1993), and indirectly found in sauropod dinosaurs as well
(Schwarz et al., 2007c). In vivo, the zygapophyseal ar-
ticular capsule is composed of fibrous connective tissue
enclosing a synovial joint between the prezygapophysis
of a vertebra and the postzygapophysis of the preceding
vertebra.

Other structures that probably acted as intervertebral
connections run along the dorsal margin of the neural

Fig. 124 - Articulated 9 and 10° cervical neural arches of Scipionyx samniticus (A), and close-up of the remnants of the right zygapo-
physeal articular capsule (B). Scale bar = 0.2 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 124 - Archi neurali articolati della 9° e 10° vertebra cervicale di Scipionyx samniticus (A), e particolare dei resti della capsula
articolare della zigapofisi destra (B). Scala metrica = 0,2 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 1 3 7

spines of the dorsal vertebrae 9, 11, 12 and 13, as well
as from the apex of the spines to the beak-like ligament
attachments described in the Osteology section (see Dor-
sal Vertebrae). Given the link to the latter, we regard
these structures as remnants of the interspinal ligaments,
whereas the most dorsally placed, apical fragments would
derive from the supraspinal ligaments. Only very thin
stripes adjacent to the bones are preserved of these liga-
ments; these are scarcely perceptible under visible light
but well-contrasted under ultraviolet light (Fig. 125).

Fig. 125 - Ultraviolet-induced fluorescence photograph of the caudal
dorsal neural spines of Scipionyx samniticus, showing possible frag-
ments of the supraspinal ligaments (arrows).

Fig. 125 - Scipionyx samniticus. Fotografia in fluorescenza indotta da
luce ultravioletta delle spine neurali dorsali caudali, che mostra i proba-
bili frammenti dei legamenti sopraspinali (frecce).

Axial articular cartilages

The Wood's lamp highlights a frankly light and bright
colouring on some vertebrae of Scipionyx, indicating the
possible presence of organic remnants in correspondence
to the neurocentral articular surfaces. More precisely, the
fluorescence runs along the articular surfaces of the dorsal
centra 6 and 10, sacral centra 1 and 5, and caudal centra 1
and 2 (e.g., Fig. 126). Given their precise localisation, we
regard these organic films as remnants of the neurocentral
cartilages, which in a very immature specimen, such as

Fig. 126 - Remains of the neurocentral articular cartilages on the 6%
dorsal (A) and 5® sacral (B) vertebrae of Scipionyx samniticus fluo-
resce under ultraviolet light. See Appendix 1 or cover flaps for abbre-
viations.

Fig. 126 - In luce ultravioletta, sulla 6° vertebra dorsale (A) e sulla
5? vertebra sacrale (B) di Scipionyx samniticus assumono fluorescenza
i resti delle cartilagini articolari neurocentrali. Vedi Appendice 1 o
risvolti di copertina per le abbreviazioni.

Scipionyx, are expected still to be abundantly present. In
fact, we remark that the neurocentral vertebral joints are
synchondroses, i.e., a type of cartilaginous joint that is
temporary, existing only during the growing phase — they
become progressively thinner during skeletal maturation
and, ultimately, become obliterated by bone union. Histo-
logically, synchondroses consist of hyaline cartilage.

Appendicular articular cartilages

UV light analysis also reveals that most of the limb
and girdle joints of Scipionyx is marked by a golden-
white film of organic remnants delimiting, with preci-
sion and continuity, the articular ends of the bones. This
film corresponds to the position of the hyaline and fi-
brous symphyseal articular cartilages (Figs. 127-128).
Articular fibrocartilaginous caps — present on the epi-
physes of the long bones of all extant archosaurs — rarely
fossilise (e.g., Geist & Jones, 1996; Reisz et al., 2005;
Schwarz et al., 2007b), so the remarkable level of pres-
ervation of Scipionyx reveals that, in the forelimbs, these
caps are more extended in the shoulder, elbow and wrist
joints, but also clearly visible even where sandwiched
in-between smaller distal elements, like between meta-
carpals and phalanges, and between proximal and distal
phalanges. In the pelvic girdle, the proximal articular

Fig. 127 - Close-ups of the remains of the articular cartilages of the
right humerus (A) and left manus (B) of Scipionyx samniticus under
ultraviolet-induced fluorescence. See Appendix 1 or cover flaps for
abbreviations.

Fig. 127 - Scipionyx samniticus. Particolare dei resti delle cartilagini
articolari dell’omero destro (A) e della mano sinistra (B), fotografati in
luce ultravioletta. Vedi Appendice 1 o risvolti di copertina per le abbre-
viazioni.

138 CRISTIANO DAL SASSO & SIMONE MAGANUCO

surfaces of the tibiae and the fibulae are evidently coated
by cartilage, as are the distal condyles of the femora, the
extremity of the pubic peduncle of the right ilium, and,
in front of it, the iliac process of the right pubis. The
ventral extremities of the pubic feet are also covered by
cartilage which, under UV light, appears indistinguish-
able in the fossil from the other areas of cartilage, even
if the cartilage of ileo-pubic and interpubic joints should
be of a fibrous nature, i.e., the type of tissue that, in ver-
tebrates, connects the symphyses, making them capable
of supporting both compressive and tractional stresses.
In order to avoid invasive analyses (e.g., drilling, thin
sectioning), which would have damaged this unique
specimen, no details of the growth zone overlain by the
articular caps have been studied in Scipionyx.

A film that fluoresces like cartilage under UV light is
seen as well on the thin caudal surface of the iliac blade
(Fig. 128). However, its position and well-delimited ex-
tension are consistent with the posterior sacroiliac liga-
ment attachment, which in vivo would have been com-
posed of fibrous connective tissue.

Muscles, connective tissue and other soft tissues
in the neck

As already noted, tissue remains are preserved in
several anatomical regions of the hatchling Italian thero-
pod. The most cranial area (and also one of the most
extended) in which these tissues can be clearly seen is
at the base of the neck, where they are observable even
with the naked eye under visible light on account of their
yellowish colour. Here, the remains occupy a flat, trian-
gular area, 19.3 mm long and 12.6 mm wide, positioned
ventral to the last cervical vertebra and the first four dor-
sals, and delimited caudally by the scapular girdle ele-
ments (Fig. 129).

In the former description of Scipionyx, the tissue in
this area was improperly referred to the pectoral muscula-
ture (pem in Dal Sasso & Signore, 1998a: fig. 2). Howev-
er, most of the original structural arrangement of the soft
tissues is lost here, so much so that the background mass,
which has a consistency intermediate between fossilised
bone and sediment, appears as an amorphous matrix. At
first sight, this area was thought to possibly include part
of the dinosaur’s skin; in actual fact, patches of somatic
musculature (Fig. 130), as well as frayed collagen bundles
(Fig. 131), are discernible within its mass under the light
microscope at a magnification of more than 50X, and, re-
markably, a tract of trachea protrudes from it as well (see
below). In extant archosaurs there is not a large amount
of connective tissue in this region, and there should be no
fat in a young hatchling, so this amorphous mass likely
derives from another kind of tissue, which suffered much
more decay. The most prevalent tissue/organ in the ven-
tral region of the base of the neck in a neonate archosaur
would be the thymus gland. In neonate crocodiles and
most neonate birds, the thymus stretches from the cra-
nial part of the lungs along the ventrolateral region of the
whole neck (Huchzermeyer, pers. comm., 2010). Never-
theless, in the absence of morphological evidence of fos-
silised Iymphoid tissue in Scipionyx, the conclusion that
this mass represents the remains of the dinosaur’s thymus
gland is a speculation based on anatomical position.

Fig. 128 - The remains of the articular cartilages of the pelvic and hind-
limb bones of Scipionyx samniticus under ultraviolet-induced fluores-
cence. See Appendix 1 or cover flaps for abbreviations.

Fig. 128 - Scipionyx samniticus. I resti delle cartilagini articolari
delle ossa del cinto pelvico e degli arti posteriori, fotografati in luce
ultravioletta. Vedi Appendice 1 o risvolti di copertina per le abbre-
viazioni.

The most evident patch of musculature in the neck
of Scipionyx runs along the ventral side of the centra of
the dorsal vertebrae 2, 3 and 4. Worth of mention is the
fact that at least two series of dark parallel bands, regu-
larly spaced by lighter intervals, perpendicular to the
craniocaudal orientation of the bundle, are visible in the
remnants of this muscle, just below the ventral margin
of the 3" dorsal centrum, at 60X magnification under
the light microscope. This banding does not match any

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 139

Fig. 129 - Soft tissues preserved at the base of the neck of Scipionyx samniticus. A) overall view in visible light. Regarding the dis-
located centrum of the 10° cervical vertebra, see Trachea in Respiratory Physiology. B) UV close-up of a brightly fluorescing strip
of tissue marking the position of the best preserved hypaxial muscle remains. Scale bar = 2 mm. See Appendix 1 or cover flaps for

abbreviations.

Fig. 129 - Tessuti molli conservati alla base del collo di Scipionyx samniticus. A) vista generale in luce visibile (in merito al centro
dislocato della 10° vertebra cervicale, vedi Trachea in Respiratory Physiology); B) particolare in luce UV di una striscia di tessuto a
fluorescenza brillante, che marca la posizione dei resti di muscolatura ipoassiale meglio conservati. Scala metrica = 2 mm. Vedi Appen-

dice 1 o risvolti di copertina per le abbreviazioni.

Fig. 130 - Close-up of the muscle remains of Scipionyx samniticus
shown in Fig. 129. The arrow points to the chromatic bands shown
also in Fig. 133. Scale bar = 1 mm. See Appendix 1 or cover flaps for
abbreviations.

Fig. 130 - Scipionyx samniticus. Particolare dei resti muscolari mostrati
in Fig. 129. La freccia indica le bande cromatiche illustrate anche in
Fig. 133. Scala metrica = 1 mm. Vedi Appendice 1 o risvolti di coper-
tina per le abbreviazioni.

relief, but is fairly chromatic in nature, with a frequency
of 24-28 dark and light bands per mm. Such an order of
magnitude is not at all consistent with the nanometric
banded structure of collagen (e.g., Kadler et al., 1996;
Lingham-Soliar & Wesley-Smith, 2008; Smith, 1968),
and it also rules out any possible artefact of prepara-
tion. More importantly, the appearance of clear banding
leads one to think of highly ordered and repeated, mi-
crometric biological structures, which, if aligned with

Fig. 131 - Possible frayed collagen bundles embedded in the amorphous
indeterminate tissue at the base of the neck of Scipionyx samniticus.
Scale bar = 1 mm.

Fig. 131 - Probabili fasci sfilacciati di collagene, immersi nel tessuto
amorfo indeterminato alla base del collo di Scipionyx samniticus. Scale
bar= 1 mm.

one another, would produce an overall striated effect
visible even under a light microscope. This phenome-
non does not occur in collagen, but is typical of somatic
skeletal musculature (e.g., Kardong, 1997). However,
the striated pattern of skeletal muscles is usually vis-
ible only with polarised light under the light microscope
(Mascarello, pers. comm., 2009) and has a micrometric
order of magnitude, inconsistent with the one we de-
scribe above.

140 CRISTIANO DAL SASSO & SIMONE MAGANUCO

In order to investigate these structures more deeply, we
took a microsample nearby. Morphological and chemical
elemental analysis of this sample under SEM confirmed
uneduivocally the presence of phosphatised muscular tis-
sue with still intact bundles of myofibres (Fig. 132A-C).
The myofibres are preserved in three dimensions, are still
appressed to one another, are parallel and straight, and
have margins that are clearly delimited by narrow spaces
left by the dissolution of the muscle cell membrane (sar-
colemma) and of the connective tissue surrounding it (en-
domysium). Remarkably, the myofibres have an internal,
continuous and regular transverse banding pattern with a
periodicity (2.9 um) consistent with the mean sarcomeric
size of extant vertebrates (Panchangam et al., 2008; Botte
& Pelagalli, 1982; Cavitt et al., 2004; Cross et al., 1981;
Mubhl et al., 1976; Wheeler & Koohmaraie, 1994). These
structures definitely differ from those produced by col-
lagen fibres, which are typically arranged in cross-striated
fibrils (D-banding) with a specific axial periodicity of 67
nm (Kadler et a/., 1996; Lingham-Soliar & Wesley-Smith,
2008; Smith, 1968). The SEM images of other samples
(Fig. 132D) show that most of the surrounding muscu-
lature has lost some detail during fossilisation: the indi-
vidual fibres can be resolved but the sarcomere-related
banding is faint or lost. Nevertheless, their arrangement in
multiple layers of packed, straight elements with constant
transverse diameter, as well as their order of magnitude,
are consistent with the morphology of fossilised verte-
brate myofibres (e.g., Chin et al., 2003; Kellner, 1996a;
Wilby & Briggs, 1997). As better explained below (see
Soft Tissue Taphonomy), an intermediate microfabric
(sensu Wilby & Briggs, 1997) might have been produced
in this area during early diagenetic stages, with muscle

fibres being replaced by apatite crystals in the form of
coarse to large agglomerations.

The biological structure responsible for the banded
pattern visible with non-polarised light under the optical
microscope (Fig. 133) remains unknown, as we have not
yet found anything comparable to that order of magni-
tude, at least in the palaeontological literature devoted to
soft-tissue preservation.

Even though we did not have the possibilty of tak-
ing samples for SEM analysis in other areas of the soft
tissues at the base of the neck of Scipionyx, we reason-
ably refer the brown-reddish fibrous structures surfacing
close to the distal half of the 9 right cervical rib (Fig.
131) as collagen fibres rather than muscle remains. This
attribution relies on the observation that Scipionyx $ my-
ofibres invariably show a clearly different aspect under
the optical and scanning-electron microscopes, i.e., a
packed, cord-like arrangement, with diameters that re-
main constant even for relatively long distances (see
images of other muscle remnants of the specimen in the
following chapters). On the contrary, despite displaying
a filamentous aspect and a craniocaudal co-orientation,
the fibrous structures close to the 9! right cervical rib
follow a curved, sinuous path, and have intersections or
V-shaped bifurcations as well; most importantly, they
show a quite variable, non constant diameter (8-40 um),
not only along a single filament, but also from fibre to
fibre. Given the micrometric order of size, which is not
consistent with the nanometric order of the single fibril,
Mascarello (pers. comm., 2009) hypothesised that each
individual filament may correspond to collagen bundles
and the bifurcations may match inhomogenous partitions
of fibrils (i.e., frayed collagen bundles).

Cumor=8.605 kV 26 cnt ID=Lulbd Zn ka2 Re la2
Window

Vert=1000

ao .p- dts sig ;
Pi te GATE
a i a È

0.005 - 40.955= 41653 cnt

Fig. 132 - Soft tissue preservation at the cellular level in the neck muscles of Scipionyx samniticus. A) SEM image of a bundle of
myofibres showing sarcomere-related banding; B) SEM element microanalysis of the same sample, indicating that the myofibres are
lithified in calcium phosphate. In other bundles of myofibres, the sarcomere-related banding is faint (C) or lost (D). See Appendix 1

or cover flaps for abbreviations.

Fig. 132 - Conservazione dei tessuti molli a livello cellulare mei muscoli del collo di Scipionyx samniticus. A) immagine SEM di un
fascio di miofibre che mostrano bande riferibili a sarcomeri; B) la microanalisi degli elementi al SEM sullo stesso campione indica che
le miofibre sono litificate in fosfato di calcio. In altri fasci di miofibre le bande dei sarcomeri sono meno marcate (C) o sono andate
perse (D). Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 141

Fig. 133 - Close-up of Fig. 130, showing parallel, regularly-spaced
chromatic bands (arrow), visible under the optic microscope and non-
polarised light in the soft tissue remains at the base of the neck of
Scipionyx samniticus. This banded pattern is maybe related to a supra-
cellular level of organisation of the muscular tissue. Scale bar = 100
um. See Appendix 1 or cover flaps for abbreviations.

Fig. 133 - Scipionyx samniticus. Particolare di Fig. 130 che mostra
bande cromatiche parallele, spaziate in modo regolare (freccia), visi-
bili al microscopio ottico e in luce non polarizzata nei resti dei tessuti
molli alla base del collo. Questa struttura a bande dipende forse da un
livello di organizzazione supracellulare del tessuto muscolare. Scala
metrica = 100 um. Vedi Appendice 1 o risvolti di copertina per le
abbreviazioni.

About 1 mm from these filaments, towards the ven-
tral edge of the amorphous soft-tissue mass, another se-
ries of parallel and regular chromatic bands can be seen.
Here, the banding seems to have a smaller size and/or a
higher periodicity (around 40 per mm), but its aspect is
quite analogous to that described above. Again, we sug-
gest that these bands support the presence of muscular
remains, but we are unable to explain the physical nature
of this optical effect.

The patches of musculature preserved at the base of
the neck of Scipionyx are too fragmentary to infer their
arrangement, their origin or their insertion sites. Neverthe-
less, we can reasonably suppose that they pertain to the hy-
pobranchial and hypaxial complexes. The latter includes,
for example, the intrinsic and extrinsic tracheal muscles,
as well as the M. sternohyoideus and the M. sternotra-
chealis (Berger, 1960; George & Berger, 1966). The M.
sternotrachealis is particularly well-developed in extant
long-necked avian theropods (Botte & Pelagalli, 1982).

Evident rows of parallel striations, with a periodicity
of at least 14-16 per mm, can be seen at the centre of
the amorphous mass at the base of the neck and in the
corner formed by the right scapular acromion and the
furcula (Fig.134A). Ata first glance, based on their mac-
roscopic aspect and size, the striations seem to represent
fascicula of myofibres, but after careful examination
under the optic microscope, these striations are seen to
disappear almost completely when the light is oriented
parallel to them (Fig. 134B). Therefore, they consist of
fine scratches (carvings) made on the soft tissue remains.
The complete lack of pigmentation confirms that these
striations are not organic traces, but artefacts of prepara-
tion; namely, they are very fine, light scratches caused
by the tiniest metal pins that were used to manually re-
move the sediment from the fossil (Rampinelli, pers.
comm., 2009).

Fig. 134 - Soft tissue remains at the base of the neck of Scipionyx sam-
niticus. Parallel striations (A, arrows) that disappear when the light is
oriented parallel to them (B) do not indicate muscle bundle orientation,
but are artefacts of preparation (fine scratches). Scale bar = 1 mm. See
Appendix 1 or cover flaps for abbreviations.

Fig. 134 - Resti di tessuti molli alla base del collo di Scipionyx sam-
niticus. Strie parallele (A, frecce) che scompaiono quando la luce viene
orientata parallelamente ad esse (B) non indicano la disposizione di
fasci muscolari ma sono artefatti di preparazione (incisioni sottili).
Scala metrica = 1 mm. Vedi Appendice 1 o risvolti di copertina per le
abbreviazioni.

Trachea

Within the triangular-shaped area of neck muscle and
connective remains, are 8-10 light-coloured, almost di-
aphanous rings, partly embedded in the putative connec-
tive tissue (Fig. 135). These structures form an about 7
mm-long row, aligned in a craniocaudal direction: cau-
dally to the furcula, they rest underneath the right cora-
coid; cranially to the furcula, they fade into the mass of
neck tissue. Therefore, these rings lie exactly where the
trachea is expected to be, i.e., in the prethoracic region,
aligned with the plane of symmetry of the bones of the
pectoral girdle.

The individual rings are found in place, one in front
of the other, and regularly separated by gaps which
measure about half the craniocaudal length of each
ring. This structure is commonly found in extant ver-
tebrates: it reinforces the ventral and lateral sides of
the trachea, and protects and maintains the airway pat-

142 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 135 - The visible tract of the trachea of Scipionyx samniticus.
Arrows point to the two tracheal rings exposed in cranial view. Scale
bar = 1 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 135 - Il tratto visible della trachea di Scipionyx samniticus. Le
frecce indicano i due anelli tracheali esposti in norma craniale. Scala
metrica = 1 mm. Vedi Appendice 1 o risvolti di copertina per le abbre-
viazioni.

ent while giving the tracheal tube maximum flexibility
(e.g., Romer & Parsons, 1977). In Scipionyx the size
of each ring is around 0.33 mm and the gaps are about
0.17 mm, so 2 rings and 2 gaps are about 1mm long. It
is difficult to measure the height of the rings, as most
of them are incomplete. Two elements (5° and 6"), oc-
cupying a central position within the row cranial to the
furcula, are exposed laterodorsally and form an incom-
plete ring opened dorsally and having dorsally rounded
apexes (in this view, both right and left apexes are vis-
ible). The lateral walls seem slightly compressed, so
these two rings have a U-shape. The height of these
rings is a little less than Imm.

The potentially best preserved ones are the two tra-
cheal rings caudal to the furcula: they are still coated
with a thin patch of ?connective tissue. The ring lying
closer to the furcula, well-exposed in cranial view, is C-
shaped, confirming that the cranialmost tracheal rings
are a bit crushed and dorsally opened. Observed in vivo,
this morphology is consistent with that of the young in-
dividuals of many extant terrestrial vertebrates and the
adults of some, in which the dorsal arches of the trache-
al rings, made of hyaline cartilage, are not chondrified
(e.g., Romer & Parsons, 1977). In Homo sapiens, for
instance, each tracheal element forms an incomplete
ring occupying the ventral 2/3 of the circumference of
the trachea; they are incomplete behind, where the tube
is completed by fibrous tissue and smooth muscle fibres
(e.g., Saladin, 2010). Incomplete chondrification (and
consequently, fossil preservation) may well explain the
relatively small size of the tracheal rings in Scipionyx.
Compared with crocodiles and birds, one would expect
an animal with the size of Scipionyx to have complete
tracheal rings with a larger diameter. For example, a 29
cm-long Nile crocodile hatchling, studied for the pur-

pose of this monograph, had a tracheal diameter at mid-
length of the trachea of 1.4 mm (Huchzermeyer, pers.
comm., 2010): this means that an individual the size of
Scipionyx would have had a tracheal diameter of about
2 mm. Nonetheless, the size of the preserved (chon-
drified) portion of the rings in Scipionyx does not cor-
respond to the actual tracheal diameter: the C-shaped
rings need almost another 1 mm to be completed, thus
reaching an overall diameter of 2 mm. We do not be-
lieve that the rings were subjected to shrinkage during
their fossilisation, nor to belong to one of the (much
smaller) animals preyed upon by Scipionyx (see Gut
Contents And Feeding Chronology).

Oesophagus and stomach

The oesophagus and stomach of Scipionyx are not
preserved; however, we have some indirect evidence
of their position. A faint trace of the oesophagus can be
seen just dorsal to the trachea (Fig. 135). About 5 mm-
long, this trace is marked by tiny fragments of bones
and scales which parallel the trachea in a craniocaudal
direction, at the level of the acromion of the right scap-
ula. This arrangement is consistent with the position
of the oesophageal tube, as seen in extant vertebrates
(e.g., Romer & Parsons, 1977). The intimate embed-
ding of the remains within the decayed tissue of the
neck rules out the possibility of them being the bodies
of organisms that were deposited of their own accord
above or below the dinosaur’s carcass. The nature of
these remains, which, consequently, are swallowed,
partially crushed prey, is discussed in a dedicated sec-
tion (Gut Contents And Feeding Chronology).

Food remains pertaining to the stomach can also be
discerned, found in the form of a cluster of tiny allog-
enous bones (for a detailed description, see Stomach
Contents). Their position indicates that the organ, or
better, the portion of it that had contained them, was
situated in the thorax of Scipionyx at the level of the
9th dorsal vertebra, close to the cranialmost tract of the
intestine (Fig. 117). Unlike what is observed for the
oesophagus, the stomach contents are not intimately
embedded in the surrounding soft tissue. Nevertheless,
the little mass of bones is certainly encased in the tho-
racic cavity, being positioned in an intermediate plane
between the 6 and 7° left and right dorsal ribs, at the
same depth as the adjacent vertebral centrum of D9.
Most importantly, the swallowed bones are partially
overlapped by the cranial portion of the duodenum:
this indicates that they are positioned in the left side
of the thorax, i.e., just where in extant archosaurs the
stomach is situated (Huchzermeyer, 2003; McLelland,
1990).

Contrary to what is seen in the intestine, the stom-
ach is not preserved even as a mould. There is reason-
able ground to suppose that the dissolution of this or-
gan, which in vivo is endowed with a thick muscular
wall, was caused by its own physiology. The vertebrate
stomach is typically a site of chemical digestive secre-
tions, collectively called gastric juice, the presence of
which has been proven indirectly also in theropod di-
nosaurs (e.g., Varricchio, 2001). Gastric juice includes
some enzymes and mucus, but is primarily composed of

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 143

hydrochloric acid released from glands in the mucosal
wall of the stomach (Kardong, 1997); in crocodiles, for
instance, the acid is capable of dropping the gastric pH
to as low as 1.2 (Huchzermeyer, 2003). Probably, the
acidic environment, controlled by physiological proc-
esses, persisted for a while even after Scipionyx died,
simply because the stomach is made to be an efficient
liquid-storage bag. Under these conditions, a strongly
acidic pH would have sped up the decay of the stom-
ach compared with the rest of the carcass. This hypoth-
esis is endorsed by the experimental observation that
an acidic environment not only inhibits bacterial ac-
tion, but also enhances the decomposition of biologi-
cal tissues (Lyman, 1994). In other terms, the physi-
ological pH present in the stomach at death prevented
the precipitation of phosphates and carbonates for long
enough to let the organ decompose (see also Soft Tis-
sue Taphonomy).

On the other hand, the intestine of Scipionyx took
more time to decay and was partially fossilised because
its lumen was characterised by a neutral pH. In fact, it
is well known (e.g., Kardong, 1997) that the pyloric
glands, opening into the terminal tract of the stomach,
produce secretions that help neutralise the acid chyme
as it moves into the intestine. Notably, also the sphe-
nodontid reptile from Pietraroja recently described by
Evans et al. (2004) preserves some intestinal portions
but not the stomach.

Nothing can be said about the shape and size of the
stomach in Scipionyx. From gut contents reported in the
tyrannosaurid Daspletosaurus, it was supposed that the
theropod stomach was similar to the two-part stomach of
birds (Varricchio, 2001). Based upon swallowed, acid-
etched juvenile hadrosaur bones and other gut contents,
and also upon tooth-marked bones, Varricchio (2001) in-
ferred that Daspletosaurus and most theropods ingested
and digested prey in a manner similarto that of extant ar-
chosaurs (crocodilians and birds), employing a two-part
stomach with an enzyme-producing proventriculus fol-
lowed by a thick-walled muscular gizzard. However, if
one neglects the presence of a tiny pyloric chamber, the
crocodilian stomach has only one chamber, lined entire-
ly by secretory glands in the mucosa, which in any case
extend into the pyloric chamber as well (Huchzermeyer,
pers. comm., 2010). Therefore, there is not the func-
tional distincetion in crocodilians seen in the stomach of
birds, in which the secretory glands are limited to the
proventriculus (which, with the exception of the ostrich
[Huchzermeyer, 2000], is small), whereas the gizzard is
lined by koilin excreted by shallow glands and serves
as storage (e.g., in raptors) or as a grinding organ (e.g.,
in the domestic fowl and in the ostrich). The fact that
pieces of undigested bone are found in the intestine of
Scipionyx (see Intestinal Contents) would suggest that
it had a bird-like, two-chambered stomach, rather than
a crocodilian stomach (in which bones are usually dis-
solved entirely). However, digestive physiology varies
considerably even in the same individual, depending on
a number of factors that do not leave any fossil trace (see
Digestive Physiology).

Summing up, in absence of a fossil stomach, in our
opinion it can be only postulated that Daspletosaurus or
Scipionyx, or any other non-avian theropod, must have
had a two-chambered stomach.

Dorsal epaxial muscles

In between the neural spines of the dorsal vertebrae
6 and 7 is a small patch of yellowish soft tissue (Fig.
136A) that markedly fluoresces a gold colour under UV
light (Fig. 136B), similar to the musculature preserved at
the base of the tail (see below). A cranial portion, next to
the distal apex of the right transverse process of the 6"
dorsal vertebra, and a caudal portion, in a more medial
position, overlaying the right postzygapophysis of the
6% and part of the right prezygapophysis of the 7! dor-
sal vertebra, are discernible. Given their position, these
soft-tissue remains likely represent part of the epaxial
musculature and epaxial connective tissue, in particular
a residual patch of the M. transversospinalis group or of
the M. longissimus dorsi.

Fig. 136 - Close-up under visible light (A) and ultraviolet-induced fluo-
rescence (B) of a residual patch of the epaxial musculature and epaxial
connective tissue (arrows) in the dorsal tract of the vertebral column of
Scipionyx samniticus. Scale bar = 2 mm. See Appendix 1 or cover flaps
for abbreviations.

Fig. 136 - Scipionyx samniticus. Particolare in luce visibile (A) e ultra-
violetta (B) di un lembo residuale della muscolatura epiassiale e del
connettivo epiassiale (frecce) nel tratto dorsale della colonna verte-
brale. Scala metrica = 2 mm. Vedi Appendice 1 o risvolti di copertina
per le abbreviazioni.

Liver and other blood-rich organs

In the thoracic portion of the visceral cavity, just crani-
al to the intestine and between the forelimbs of Scipionyx,
a halo made by a reddish matter is seen to impregnate the
sediment, as well as the left arm elements, the cranial gas-
tralia and some dorsal ribs (Figs. 137A, 138A-C). Basing
their hypothesis on colour and anatomical position, Dal
Sasso & Signore (1998a) postulated in the former descrip-
tion of Scipionyx that this reddish macula was composed
of iron-rich minerals derived from the decay of the liver,
which in vertebrates is the single largest visceral organ
(e.g., Schaffner, 1998) and the major organ accumulating

144 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 137 - A) the reddish halo deriving from the decay ofthe liver (and of other blood-rich organs) of Scipionyx samniticus, seen under
visible light; B) ultraviolet-induced fluorescence reveals that this material impregnates a larger area of the thorax. The distinct blue-
indigo fluorescence corresponds to the primary light emission peak of biliverdin and, therefore, might be consistent with the presence
of bile pigment residues. Scale bar = 5 mm.

Fig. 137 - Scipionyx samniticus. A) l’alone rossastro derivante dalla decomposizione del fegato (e di altri organi ricchi di sangue),
fotografato in luce visibile; B) una foto in luce ultravioletta mostra che questo materiale ha impregnato un’area più ampia del torace.
La ben distinta fluorescenza in blu-indaco corrisponde al picco di emissione primaria della biliverdina e dunque è compatibile con la
presenza di pigmenti residui di bile. Scala metrica = 5 mm.

Fig. 138 - Close-ups taken under the optical microscope, with magnification increasing from A to B to C, of the reddish material
encrusting bone and sediment in the ventral portion of the thorax of Scipionyx samniticus. Scale bars = 1 mm. See Appendix 1 or cover
flaps for abbreviation.

Fig. 138 - Particolare al microscopio ottico, con ingrandimento crescente da A a B a C, del materiale rossastro che incrosta le ossa e il
sedimento nella porzione ventrale del torace di Scipionyx samniticus. Scale metriche = 1 mm. Vedi Appendice 1 o risvolti di copertina
per l’abbreviazione.

blood (e.g., Kardong, 1997). Under UV light (Fig. 137B),
the reddish halo becomes subcircular in outline and is
seen to impregnate a larger area of the rib cage, with a
diameter of about 17 mm. More remarkably, it produces
a distinet blue-indigo fluorescence under UV light, which
corresponds to the primary light emission peak of biliver-
din and is, therefore, consistent with the presence of bile
pigment residues (Ruben et a/., 1999).

In the absence of direct chemical evidence, the char-
acterisation of the supposed iron-rich mineral remained
non-compelling until we conducted SEM element miero-
analysis. We examined six microsamples of the red mat-
ter encrusting the bone and the matrix (see Appendix 7),
and all of them showed the same composition, unique to
that area: hydrated iron oxide (limonite). The iron peak
is as high as that of the gold coating (Fig. 139), and there

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 145

SM umore 14.665 keY Geni ID»
eri=1000 Window 0.005 - 40.955= 100896 e1

Fig. 139 - Scipionyx samniticus. SEM imaging (A) and element microa-
nalysis (B) of the reddish material shown in Fig. 138 indicate that it is
composed of microcrystals of hydrated iron oxide (arrows). Note that
the iron peak is as high as the one of the gold coating, and that there
are only traces of carbonates and phosphates, which are the prevailing
components of the sediment (Fig. 13A) and of the fossil (e.g., Figs.
121C, 132B). See Appendix 1 or cover flaps for abbreviations.

Fig. 139 - Scipionyx samniticus. Le immagini al SEM (A) e la micro-
analisi degli elementi (B) del materiale rossastro mostrato in Fig. 138
indicano che questo è composto da microcristalli di idrossido di ferro
(frecce). Si noti che il picco del ferro è alto quanto quello dell’oro
derivante dalla doratura del campione, e che vi sono solo tracce di
carbonati e fosfati, che sono i componenti prevalenti del sedimento
(Fig. 13A) e del fossile (es., Figs. 121C, 132B). Vedi Appendice 1 o
risvolti di copertina per le abbreviazioni.

are only traces of carbonates and phosphates, which are
the prevailing compounds in the matrix (Fig. 13A) and
the fossil (e.g., Figs. 121C, 132B), respectively. There-
fore, this iron cannot be allogenous — for instance deriv-
ing from the pyrite formed from sulphate reduction in the
centre of a carcass and drawing iron from the sediment as
well as from the carcass itself — because in this case one
would find the presence of its drainage into the surround-
ing environment. Rather, because of its single localisation
and large amount, this iron represents an endogenous ele-
ment derived from post mortem degradation of the dino-
saur’s haemoglobin.

This haematic nature, and its association to possi-
ble bile remains, strengthens the idea that the reddish
halo is a product of liver decay. Such an interpretation
is also consistent with the anatomical position of the
liver in extant archosaurs: in the Nile crocodile, the
liver is made from two lobes embracing the heart situ-
ated between the 4° and 8 thoracic rib (Huchzermeyer,
2003); in birds, the liver lobes embrace cranially the
caudal half of the heart, and the two organs are ventrally
protected by the large sternum (McLelland, 1990: figs.
120-121). Given the proximity of the two organs, it can
be hypothesised that the stain may derive also from the
decay of Scipionyx ’s heart, although no trace of it, in the
form of muscular tissue, is present in the specimen. Ac-
cording to Huchzermeyer (pers. comm., 2010), a third
contribution to the large reddish halo might have been
made by the spleen, another blood-storing organ that in
extant archosaurs, similar to the heart, is positioned in-
between the two liver lobes, but limited to their caudal
halves.

Intestine

The intestine is the largest, most complete and most
visible internal organ of Scipionyx: its three-dimensional
loops are lumpy and shiny, reminiscent of the aspect one
would see after dissecting a modern animal (Fig. 140).
Whereas the cranial portion is tangled around itself, the
caudal portion runs ventral and parallel to the caudal
dorsal vertebrae, suggesting that it was attached to the
roof of the abdominal cavity by a mesentery, and then
passes through the pelvic cavity between the pubic and
ischial shafts, ending at the base of the tail (Figs. 115-
117). The surface of the digestive tube appears to be
more irregular in the cranialmost portion, because of the
presence of not completely liquefied ingested food, the
nature of which is described in a dedicated section (see
Intestinal Contents).

Fig. 140 - A grazing view of the intestine of Scipionyx samniticus highlights the preservation of the organ in three dimensions.
Fig. 140 - Una vista radente dell’intestino di Scipionyx samniticus evidenzia la conservazione tridimensionale dell’organo.

146 CRISTIANO DAL SASSO & SIMONE MAGANUCO

A relevant feature is the intestine’s position: this organ
is placed much further cranially in the abdominal cav-
ity than was previously thought for dinosaurs (e.g., Paul,
1988; Wellnhofer, 1985), so, consequently, the intestine
was not supported by the pubic bones. Because soft tissues
by their very nature are quite mobile, it is always difficult
to be certain that they have retained their original position
in fossils. However, Scipionyx was very slowly and plas-
tically deformed during a long-term diagenetic process.
Besides that, the gastralia provide convincing proof that
at least some portion of the intestine is positioned as it

Fig. 141 - Close-up of the duodenal loop of Scipionyx samniticus,
showing the plicae circulares (circular folds) of the mucosa. Scale bar
=2 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 141 - Particolare dell’ansa duodenale di Scipionyx samniticus, in
cui sono visibili le plicae circulares (pieghe circolari) della mucosa.
Scala metrica = 2 mm. Vedi Appendice 1 o risvolti di copertina per le
abbreviazioni.

was in vivo. In the Pietraroja theropod, these tiny, fragile
abdominal ossifications, that are usually dispersed in the
sediment, are all very well aligned. Some gastralia follow
the curvature of a ventral loop of the intestine, indicating
that this loop has not moved with respect to the skeletal
elements. The shapes of the gastralia and of the intestinal
loop are complementary, to a point that the former seem to
embrace the latter, contributing to its support (Fig. 141).
If real, this support must be regarded as indirect, because
the gastralia cannot have been anatomically connected
to the intestine: the gastralia are part of the body wall,
strengthening it and, as such, support the intestinal mass
not individually but as a whole (see Gastralia).

A close-up of the duodenal loop (Fig. 141) shows that
the intestine is mostly (but not simply) an endocast, as its
well-exposed, sometimes anastomosed, circular folds (pli-
cae circulares) are typical of the mucosal layer, which in
vivo represents the innermost tissue of vertebrate gut (e.g.,
Kardong, 1997). Actually, remnants of longitudinal fibres
arranged perpendicular to the circular folds are seen here
and there (e.g., Fig. 142), so some patches of muscularis
externa and/or adventitia (i.e., the sheaths of smooth mus-
culature and fibrous connective tissue that in vertebrate
gut wrap around the mucosa) have been likely preserved.
As a matter of fact, our recent SEM microsamples docu-
ment the preservation of some phosphatised tissue lay-
ers, including a multi-layered epithelium in the duodenum
(Fig. 143), and capillary-sized, branched blood vessels in
the rectum (Fig. 144). This evidence clarifies once and for
all that the supposition that the intestine of Scipionyx was
a cololite (e.g., Holtz ef al., 2004) was simplistic and due
to indirect knowledge of the specimen, the preservation
of which is without doubt more complex than one could
expect (see also Soft Tissue Taphonomy).

Because the intestine of Scipionyx folds over upon
itself several times, at first sight it seems impossible to
discriminate its regions. However, based on their relative
position, shape, length, relative diameter, texture, demoli-
tion status of the contents, remnants of mesenteric con-

Fig. 142 - Close-ups, with magnification increasing from A to B, of a tract of the descending loop of the duodenum of Scipionyx sam-
niticus, with remnant longitudinal fibres of the muscularis externa and/or adventitia layer arranged perpendicular to the inner layer of
plicae circulares. Scale bars = 1 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 142 - Scipionyx samniticus. Particolari, con ingrandimento crescente da A a B, di un tratto dell’ansa discendente del duodeno con
residui delle fibre longitudinali della muscularis externa e/o della tonaca avventizia, disposte perpendicolarmente allo strato interno
delle pieghe circolari. Scala metrica = 1 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY | 47

i A A3I he Ocnt ID*
gl Perte2000 Window 0.005 - 40.955= 77690 cn

Fig. 143 - Soft tissue preservation at the cellular level in the intestine
of Scipionyx samniticus. A) SEM image of a multi-layered visceral epi-
thelium found in a microsample of the duodenal loop; B) close-up of
one of the cell layers shown in A; C) SEM element microanalysis of the
same tissue. See Appendix 1 or cover flaps for abbreviations.

Fig. 143 - Conservazione dei tessuti molli a livello cellulare nell’intestino di
Scipionyx samniticus. A) immagine SEM di un epitelio viscerale pluristratifi-
cato trovato in un microcampione dell’ansa duodenale; B) particolare di uno
degli strati di cellule mostrati in A; C) microanalisi degli elementi al SEM sullo
stesso epitelio. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

815 keV Oent ID=
'ert=2000 Window 0.005 - 40.955= 80586 cnt

r

nections and/or ligaments and comparative anatomy with
extant archosaurs, and with the help of computed tom-
ography in-sequence slicing, we propose below our most
likely morpho-functional subdivisions (Fig. 145).

Duodenum - The duodenum constitutes a widely ex-
posed region of the intestine of Scipionyx. We are confì-
dent in attributing its ventralmost protruding portion to
the duodenal loop, as it shares at least four characters with
extant archosaurs (Botte & Pelagalli, 1982; Huchzermey-
er, 2003; McLelland, 1990; Ziswiler & Farner, 1972): U-
shaped loop made by two appressed tubes; craniocaudal
direction and asymmetrical position on the right side of
the abdomen; craniodorsal proximity to liver and stom-
ach; and higher density of folds in the mucosa layer. The
proximity to the liver and stomach is an anatomical con-
straint of most extant vertebrates due to two links, the gas-
troduodenal ligament and the hepatoduodenal ligament.
Similarly, the descending loop and the ascending loop are
held together all along their length by a mesenteric con-
nection, which usually encloses also the ventral portion of
the pancreas. The position of the duodenum in Scipionyx
matches these conditions, to a point that we can reason-
ably think that mesenteries and ligaments survived the de-
cay of the carcass long enough to hold the intestinal loops
in place until they became fossilised.

The duodenum of Scipionyx has an average diameter of
5.2 mm. Its fossilisation in the form of an internal mould
is exquisite also at a microstructural level, to a point that
in the U-turn between the descending loop and the as-
cending loop, the folds of the mucosa, that measure a few
tenths of a millimeter, are clearly visible to the naked eye
(Fig. 141). Similarly to crocodilians (Dal Sasso & Maga-
nuco, pers. obs., 2009) and birds (McLelland, 1990), the
U-turn is the most ventrally situated part of the intestine in
the abdomen of Scipionyx. The tract embraced by the gas-
tralia looks like a caecum but, most likely, it is the point

C

imrsor=27.095 keV Ocnt ID=
‘ert=1000 Window 0.005 - 40.955= 54486 cnt

Fig. 144 - Soft tissue preservation at the cellular level in the intestine of Scipionyx samniticus: SEM element microanalysis (A) and SEM
imaging (B) of a phosphatised capillary-sized blood vessel in a microsample of the rectum. The vacuolar aspect of the matrix, which has
the same chemical composition (C) as the capillary, is consistent with the tissue having been replaced by pseudomorphed phosphatised
bacteria, many of which are preserved as hollow spheres (see also Fig. 170). See Appendix 1 or cover flaps for abbreviations.

Fig. 144- Conservazione dei tessuti molli a livello cellulare nell’ intestino di Scipionyx samniticus: microanalisi degli elementi alSEM(A)e
immagine SEM(B)diunvaso sanguigno dellataglia di un capillaretrovato inun microcampionedelretto. L'aspetto vacuolare della matrice,
che ha la medesima composizione chimica (C), è compatibile con una sostituzione dei tessuti da parte di batteri fosfatizzati pseudomorfi,
molti dei quali sono conservati come cavità sferiche (vedi anche Fig. 170). Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

148 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 145 - The intestine is the most complete internal organ of Scipionyx samniticus. In this false-colour image, various parts of the
intestine are distinguished according to the morphofunctional subdivisions proposed in the text. Scale bar = 5 mm. See Appendix 1 or
cover flaps for abbreviations.
Fig. 145 - L’intestino è l’organo interno più completo di Scipionyx samniticus. I cambi di colore virtuale seguono la suddivisione mor-
fofunzionale proposta nel testo. Scala metrica = 5 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

where the descending loop surfaces again after a medial
invagination. In Crocodylus, for instance, the duodenum
does not consist of a simple U-shape, but folds over again,
forming a double loop (Huchzermeyer, 2003).

The cranialmost side of the descending loop contains
two adjacent clusters of ingested organic remains, likely
consisting of horny squamae from a small lizard-like rep-
tile, and a single vertebra, possibly referred to a fish, the
description of which can be found in the Intestinal Con-
tents section. Here we point out that the abundance of un-
digested food in this region of the intestine supports the
idea that this loop may be not far from the connection to
the stomach.

Jejunum - Towards the left side ofthe abdominal cav-
ity, a portion of intestine that is more deeply embedded in
the sediment runs medial to the dorsal edge of the ascend-
ing loop of the duodenum, then continues dorsally to the
right side, overlapping the centra of the dorsal vertebrae.
We refer this portion to the jejunum, as it shares the fol-
lowing characters with most extant archosaurs (Botte &
Pelagalli, 1982; McLelland, 1990; Huchzermeyer, 2003):
supra-duodenal localisation; lower density of mucosal
folds; and relatively uncomplicated arrangement in the
form of short, garland-like coils in the dorsal portion of

the abdomen. In addition, the visible tract of the jejunum
has a smaller diameter than the duodenum, and a fewer
solid inclusions.

It is unclear what point of the digestive tract marks
the transition from duodenum to jejunum, but probably
it is where the intestine slopes caudoventrally along
the left side of the abdomen, under and parallel to the
shaft of the right dorsal rib 9. Thus, the jejunum likely
originates from a medial invagination that the ascend-
ing loop of the duodenum seems to form after reaching
back cranially the initial tract of the descending loop.
As a matter of fact, in extant avian theropods (Botte &
Pelagalli, 1982) the transition point is found right where
the ascending loop of the duodenum reaches the caudal
wall of the gizzard, first curving dorsally, then U-turning
caudally.

So, in Scipionyx the first tract of the jejunum should be
the one that runs caudoventrally under the 9° right dorsal
rib. After making an abrupt, swollen curve, of which only
the dorsocaudal margin can be seen, the jejunum disap-
pears under the duodenum, at the level of the 12° dorsal
vertebra, and appears again at the level of the 10 dorsal
vertebra, between the shafts of the right dorsal ribs 7 and
8, before heading sinuously dorsally and vanishing. Un-
der UV light, residual fluorescence of organic remnants

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 149

Fig. 146 - Remnants of the jejunum of Scipionyx samniticus under visible
light (A) and ultraviolet-induced fluorescence (B). The latter reveals that
this tract of the intestine follows a garland-like path over the dorsal verte-
brae. Scale bar = 2 mm. See Appendix 1 or cover flaps for abbreviations.
Fig. 146 - Residui del digiuno di Scipionyx samniticus, fotografati in
luce visibile (A) e in fluorescenza indotta da luce ultravioletta (B).
Quest'ultima mostra che, con un andamento a festoni, questo tratto
dell’intestino si è sovrapposto alle vertebre dorsali. Scala metrica=2 mm.
Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

impregnating the sediment and superimposed on some
vertebrae can be seen, revealing that this tract of the jeju-
num had a garland-like path (Fig. 146). This arrangement
is consistent with post mortem persistence of the jejunal
mesentery, which prevented the jejunum from moving
away from the vertebral column after death but allowed
it to overlap onto the centra of the 10° and 11 dorsal
vertebrae and the base of the neural arch of the 12", a
position that it certainly did not occupy in life. Portions of
this tract of the intestine were probably removed during
the early coarse preparation of the fossil when trying to
uncover the vertebral column.

The final tract of the jejunum is better preserved and
still clearly visible under the light microscope: this por-
tion goes back down ventrally, trapped and protected by
the adjacent facing articular faces of the 12" and 13" dor-
sal vertebral centra. Here, a tiny patch of visceral muscu-
lature has been preserved (Fig. 147A). Remnants of lon-
gitudinally arranged visceral musculature (?muscularis
externa) also lie just ventral to the 10°-11'* dorsal centra
(Fig. 147B).

The jejunum contains a single solid inclusion: a some-
what round cluster of dozens of small cylindrical elements
(see Intestinal contents) found next to the cranial face of
the 11° dorsal centrum, more or less the same size of the
latter. Hollow, elongated structures, which we regard as
remnants of the cranial mesenteric artery and vein, or
their branches (see below), are seen immediately ventral
to this cluster.

Ileum - Just cranial to the pelvic girdle, precisely at
the level of the 13! dorsal vertebral centrum, the intestinal
tube seems to form a natural, non-artefactitious constric-
tion. This constriction is reminiscent of a possible ileorec-
tal valve (Botte & Pelagalli, 1982; Kardong, 1997) and
is the main clue leading to our tentative ascription of this
portion of intestine to the ileum. The ileum is generally
known as a considerably short tract, connecting the jeju-
num to the rectum, but in crocodilians, for instance, the il-
eum is about as long as the jejunum (Huchzermeyer, pers.

Fig. 147 - Tiny patches of visceral musculature of the jejunum of Scipionyx samniticus, preserved in between adjacent dorsal vertebral
centra (A), and ventral to the preceding dorsal centra (B). Scale bars = 1 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 147 - Minuti frammenti di muscolatura viscerale del digiuno di Scipionyx samniticus, conservati fra centri vertebrali dorsali adiacenti (A)
e ventralmente a centri dorsali che li precedono (B). Scale metriche = 1 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

150 CRISTIANO DAL SASSO & SIMONE MAGANUCO

comm., 2010). Therefore, its length in Scipionyx remains
unknown, and the pelvic bones hiding it prevent us from
observing its arrangement. A couple of tiny, somewhat
round bone fragments, maybe two vertebrae are embed-
ded in the supposed ileum (see Intestinal Contents).

Rectum - Following the basic bauplan of the digestive
system in most vertebrates (e.g., Kardong, 1997), we refer
to the large intestine the terminal, enlarged tract of the gut
of Scipionyx that, after passing through the pubic and the
ischial foramina, reaches the base of the tail. According to

some authors (e.g. Ziswiler & Farner, 1972; Botte & Pela-
galli, 1982), the large intestine includes the caeca, the co-
lon, the rectum and the cloaca. Actually, in extant archo-
saurs - crocodiles and birds - the cloaca is not part of the
large intestine (the rectocoprodaeal valve is understood
as the equivalent of the anus of mammals), and there is
no distinction between colon and rectum, so that this part
of the digestive tract is referred to as the rectocolon, or
usually just as the rectum (Huchzermeyer, pers. comm.,
2010). Therefore, the rectum starts where the ileum ends
(caeca present or not), often without a clear transition.

Fig. 148 - Computed tomography of the pelvic region of Scipionyx samniticus. The images (A-D) are of a sequence of 4 parallel
parasagittal sections moving from the left (A) to the right side (D) of the specimen. Note the U-shaped tract of the rectum, obliquely
embedded in the sediment (arrows).
Fig. 148 - Tomografia computerizzata della regione pelvica di Scipionyx samniticus. La sequenza di immagini (A-D) mostra 4 fette

parasagittali parallele che tagliano l’esemplare dal suo fianco sinistro (A) al suo fianco destro (D). Notare il tratto a U del retto,
immerso obliquamente nel sedimento (frecce).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY ss

This is the case in Scipionyx, too, in which the caeca are
absent, like in crocodilians and many carnivorous and
piscivorous birds (Huchzermeyer, 2003). In fact, caeca
basically are fermentation chambers, specialisations for
herbivorous amniota.

With respect to most extant tetrapods, Scipionyx
shows a quite anomalous caudodorsal displacement of
the rectum. Notably, this displacement is remarkably
similar to the one that can be observed in Nile crocodile
hatchlings before their voluminous yolksac is resorbed
(Fig. 110; Huchzermeyer, pers. comm., 2010). In fact,
the yolksac of reptile and bird hatchlings occupies most
of the pelvic cavity, and so during the first days of life,
their intestines undergo considerable rearrangement,
moving in to fill the space left free as the yolk gets re-
sorbed — a process that takes one to three weeks in extant
archosaurs (see Ontogenetic Assessment). Occasionally,
a yolksac can persist longer, a pathological condition
called yolksac retention: this happens when the duct be-
tween the yolksac and the intestine becomes closed due
to infection (Huchzermeyer, pers. comm., 2010). The
average time of absorption of the yolksac in extant ar-
chosaurian hatchlings, and the ontogenetic assessment
of Scipionyx, are consistent with an almost full yolksac
having displaced most of the rectum caudally. Conse-
quently, by assuming the displacement of the rectum as
entirely due to the presence of a yolksac, according to
Huchzermeyer (pers. comm., 2010) the estimated age of
Scipionyx at death would be about three days, absolute
maximum one week.

Apart from that hypothesis (no vitelline diverticulum
is seen opening opposite the branches of the cranial me-
senteric artery, nor any fossilised yolksac-like sac or yolk
remnant), it appears that the rectum of Scipionyx, once
emerged from the ischial foramen, runs ventrocaudal-
ly along the ischial shafts to their extremities, and then
curves dorsally to reach a straight and short, craniocau-
dally aligned, terminal tract, that contains an apparent fae-
cal mass. Actually, CT scan imaging shows an underlying
tract, intermediate between the one contacting the ischial
shafts and the one containing the faecal mass, obliquely
embedded into the matrix in the form of a U tube (Fig.
148). In Scipionyx, the diameter of the rectum is greater
that that of the jejunum and the supposed ileum, but in
any case does not attain the diameter of the duodenum,
just like in birds (Ziswiler & Farner, 1972). The surface of
the rectum is lumpy, but cranial to the faecal mass no vis-
ible solid inclusion stands out from the amorphous, grey-
yellowish contents.

A mysterious, micrometric structure was found in a
SEM sample of the rectum (Fig. 149A). It consists of a
very thin, hemispherical sheet, densely punched with cir-
cular holes forming a sort of reticulum. Its chemical com-
position (Fig. 149B), consistent with all other remains
of the dinosaur (i.e., calcium phosphate), together with
the micrometric diameter of the holes (2-8 yum), excludes
any artefact or trace of commonly used human products.
Among biological structures, such a reticulum is remi-
niscent of some sort of absorption or filtering membrane,
and its localisation in the rectum would lead to think of
the function of liquid re-absorbtion. This might be more
likely inferred by the presence of capillary-sized blood
vessels in the same sample (Fig. 144B). On the other
hand, the very small size and the very limited extension

1)
usor=28 925 leY Ocnt ID=
Window 0.005 - 40955= S014cnl

Fig. 149 - SEM imaging (A) and SEM element microanalysis (B) of
an enigmatic micrometric structure found in the rectum of Scipionyx
samniticus. Its chemical composition is consistent with all other nearby
fossil remains, thus excluding it is a human artefact. See Appendix 1 or
cover flaps for abbreviations.

Fig. 149 - Immagine al SEM (A), e relativa microanalisi degli ele-
menti (B), di una enigmatica struttura micrometrica trovata nel retto
di Scipio-nyx samniticus. La sua composizione chimica è analoga a
quella di tutti gli altri resti fossili circostanti, per cui si deve escludere
che si tratti di un artefatto. Vedi Appendice 1 o risvolti di copertina per
le abbreviazioni.

of this membrane with respect to the sample suggests that
it might be part of an allogenous, ingested and, maybe,
partially digested object (?plant cuticle, ?animal integu-
ment). In the palaeontological literature, we found just
one similar SEM image (Zhang et al., 2010: fig. lc), ob-
tained from an isolated pennaceous feather dated back to
the Early Cretaceous of Inner Mongolia, China. Follow-
ing Zhang ef al. (2010), the reticulum of our microsam-
ple might be degraded (originally keratinous) feather ma-
trix, and the holes within it might represent melanosome
moulds. However, some parameters do not fit the melano-
some hypothesis: the holes in the sample of Scipionyx
have a larger order of magnitude (about 5:1) and have a
lower density than the mouldic melanosomes illustrated
by Zhang et al. (2010), and the margins of the isolated
hemisperical membrane do not show any anastomosing
ridge suggesting the presence of repeated adjacent units.
At present, the lack of a truly comparable biological struc-
ture in the SEM images we examined does not allow us to
give a more reliable interpretation.

Faecal pellet - The rectum of Scipionyx terminates just
in front of a compact faecal mass, more properly referred
to as a faecal pellet, that embeds a cluster of bright scaly
remains, aligned in a craniocaudal direction for about 6
mm (see Intestinal Contents). With the aim of bringing
to light these remains and determining their nature, some
work was done in this area that changed the original ap-
pearance of this tract of the intestine (Fig. 150A); actually,
the oldest photos of the fossil (Fig. 150B) document that
a thin, opaque sheet of ochre-yellowish material covered
the scaly remains, embedding them in an internal mould
of the gut similarly to the preceding portions.

152 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 150 - The rectum of Scipionyx samniticus shows an abrupt shift from wide loops to a narrow, straight terminal tract (A). Here, the
gut moulds itself around a faecal mass containing many scales. This mass was visible even before preparing the specimen in detail (B).
Scale bar = 2 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 150 - Il retto di Scipionyx samniticus presenta un punto di brusca transizione, passando da anse voluminose ad un tratto termi-
nale stretto e rettilineo (A). Qui, anche prima che l’esemplare venisse preparato in dettaglio (B), affiora un modello interno delle
interiora fatto di resti scagliosi immersi in una massa fecale. Scala metrica = 2 mm. Vedi Appendice 1 o risvolti di copertina per le
abbreviazioni.

At least four characteristics lead us to think that this
tract of the intestine of Scipionyx is actually the terminal
tract of the rectum rather than the cloaca hypothesised in
previous works (Dal Sasso, 2001, 2003, 2004), or the co-
prodaeum, which was mistakenly identified in crocodiles
by Kuchel & Franklin (2000) as the anatomical continuum
of the rectum (Huchzermeyer, pers. comm., 2010): as men-
tioned above, this tract of the intestine is straight, forming
an abrupt restriction and an abrupt angle with the preceding
loops; its cranial half emerges from the left side ofthe body,
overlapped by the preceding loops; it is nested in a frankly
internal position, close to the vertebral column and flanked
all along its ventral side by caudofemoral musculature; and
it is the site of storage of a compact faecal mass.

Its straightness, as the Latin name rectum indicates,
represents the commonest morphological aspect in the di-
gestive system of vertebrates, and, therefore, a good di-
agnostic character for the end of the large intestine. Even
in the fossil Scipionyx, a straight arrangement clearly dis-
tinguishes the terminal tract of the gut from the preceding
ones. With regard to its position on the left side, such a
localisation can be observed in the terminal portion of the
rectum of some extant archosaurs: in Nile crocodiles, for
example, a short rectum emerges caudally from the left
side of the body, stretching medially only when it reaches
the cloacal chambers (Huchzermeyer, pers. comm., 2010).
As for the nearby musculature, the faecal pellet seems to

be trapped in-between two muscle bundles, which actually
do not have any relationship with it (see Caudofemoral
Muscles). In fact, these bundles are unlikely referred to
rectal or cloacal muscles because their myofibres are defi-
nitely straight and craniocaudally directed. This feature is
not consistent with the circular pattern of sphincters, which
are unpaired muscles located ventrally along the plane of
symmetry of the body. Attribution to highly specialised,
copulation-related muscles (e.g., anal sac constrictors)
associated to the cloaca of some reptiles (e.g., Kardong,
1997: fig. 14.43) is unlikely because of the immaturity of
our specimen, and is misleadingly suggested by the abrupt
caudal termination of the two muscle bundles (such a ter-
mination in Scipionyx is an artefact of preparation — see
Caudofemoral Muscles). In any case, the position of a pur-
ported cloacal opening would have been more ventral (pe-
ripheral) both in the fossil and in vivo (see Reconstructing
Scipionyx): osteological markers of the position possibly
reached by the cloacal opening are the ischial feet, as they
are the site of attachment of the ventral base of the tail. The
faecal pellet of Scipionyx is rather far from the ischial feet,
thus it cannot be interpreted as cloacal contents. Moreo-
ver, in extant vertebrates that possess a cloaca (including
all reptiles and birds), in normal conditions the faeces ac-
cumulate in the rectum, whereas the coprodaeum and the
more distal cloacal chambers are a transit-way out of the
body (e.g., Skoczylas, 1978; Ziswiler & Farner, 1972).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 153

Fig. 151 - This photograph of the abdomen of Scipionyx samniticus,
taken prior to preparation, documents the untouched blood vessels
(arrows) of the cranial mesenteric artery. Scale bar = 2 mm. See Appen-
dix 1 or cover flaps for abbreviations.

Fig. 151 - Questa fotografia dell’addome di Scipionyx samniticus, scat-
tata prima della preparazione, documenta l’integrità dei vasi sanguigni
dell’arteria mesenterica craniale (frecce). Scala metrica = 2 mm. Vedi
Appendice 1 o risvolti di copertina per le abbreviazioni.

In extant crocodiles, the faecal pellet is formed in the
very short rectum (Huchzermeyer, pers. comm., 2010).
Scipionyx seems to fit this pattern. In crocodiles, the recto-
coprodaeal valve connects the rectum to the coprodaeum,
which together with the urodaeum forms the urine cham-
ber of the cloaca. Crocodiles, as well as ostrich and rhea,
excrete urine separately from the faeces. The mixture of
urine and faeces in the rectum of flying birds appears to
be a later adaptation (Huchzermeyer, pers. comm., 2010).
Internal retention and compactness of the faecal pellet of
Scipionyx suggests that, like extant crocodiles and ratites,
theropod dinosaurs did possess a rectocoprodaeal valve,
and did not mix urine and faeces. Internal retention and
compactness of the faeces might suggest also that Scipio-
nyx did not suffer post mortem relaxation of the sphincters,
just like modern crocodiles. This possibility will remain
impossible to ascertain, because no anatomical structures
or intestinal products have been preserved caudal to the
faecal pellet described here. If fossilised, the cloacal tis-
sue and its contents were likely removed during the ear-
lier preparation, as the deeply excavated matrix caudal to
the preserved faecal pellet and cranial to the caudalmost
soft-tissue remains attest in the oldest photographs of the
specimen (Figs. 18, 150B).

Mesenteric blood vessels

A number of fine, delicate, almost uncoloured and
hollow filamentous structures, which were already vis-
ible before preparing the specimen, lie superimposed on
the surface of the jejunal loops (Fig. 151). They have a
wavy arrangement and a prevailingly dorsoventral orien-
tation (Fig. 152), with diameters that vary either along a

Fig. 152 - Close-ups of two selected portions of the abdominal area of Scipionyx samniticus illustrated in Fig. 151 (A, ascending loop
of the duodenum; B, jejunum), after preparation, highlighting the wavy arrangement and the hollowness of the fossilised blood vessels.
Scale bars = 0.5 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 152 - Scipionyx samniticus. Ingrandimenti di due porzioni dell’area addominale illustrata in Fig. 151 (A, ansa ascendente del
duodeno; B, digiuno) e fotografata dopo la preparazione dell’esemplare, che evidenziano le ondulazioni e le cavità dei vasi sanguigni
fossilizzati. Scale metriche = 0,5 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

154 CRISTIANO DAL SASSO & SIMONE MAGANUCO

single element (0.04-0.1 mm in the largest one) or from
one element to another (the thinnest one measures 0.02
mm); their preserved length does not exceed 10 mm.
These filamentous structures run obliquely, more-or-less
paralleling the first tract of the jejunum, from the ventral
edge of the 10! dorsal centrum to the caudoventral loop
of the jejunum, where the latter curves underneath the
duodenum (i.e., at the level of the caudal margin of the
12° dorsal centrum).

Hollow, filamentous structures are found in Scipio-
nyx also at the base of the tail; however, these are longer,
greater in diameter, evidently coloured, frequently bi-
furcated and imbricate caudally, and appear to be more
stiffened. Therefore, we maintain that, despite a superfi-
cial similarity, the two hollow structures have different
origins.

A frankly sinuous arrangement indicates that the hol-
low structures overlaying the intestine were elastic and
flexible. Given these features and the presence of some
branching and anastomosing, we regard these delicate,
nearly transparent structures as blood vessels, ruling
out any possible integumentary derivative, such as hair,
bristles or protofeathers. The preservation of structures
as delicate as blood vessels must not surprise too much,
especially in Scipionyx. Janvier (pers. comm., 2009),

for instance, reports that after axial musculature, blood
vessels are the most frequently found preserved sofît tis-
sue in fossil fishes.

A survey of the surface of the whole intestine con-
firmed that in Scipionyx the preservation of macroves-
sels (i.e., blood vessels that are visible under the light
microscope) is limited to the ascending loop of the du-
odenum, where they are very rare, and to the jenunum,
where they are abundant. The intimate contact between
the vessels and the dorsal loops of the jejunum, and the
caudal loop of the duodenum as well (Figs. 117, 151),
is in favour of the attribution of all preserved blood ves-
sels to the branches of the cranial mesenteric artery and
vein. In fact, such a distribution is consistent with the
cardiovascular anatomy of extant vertebrates, in which
the cranial mesenteric artery, arising from the abdomi-
nal aorta, branches into the caudal duodenal pancreatic
artery, the middle and right colic, as well as the jeju-
nal and ileocaecocolic arteries (e.g., McLelland, 1990;
Pinto e Silva ef a/., 2008). Notably, in crocodilians the
cranial mesenteric artery meets the intestine right at the
point at which the jejunum, after running straight along
the dorsal aspect of the abdominal cavity, becomes sus-
pended in loose coils by the mesentery (Huchzermeyer,
2003).

Fig. 153 - Bundles of somatic muscles are remarkably preserved in three dimensions (see also Fig. 158) at the base of the tail of Scipio-
nyx samniticus. Scale bar = 5 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 153 - Fasci di muscoli somatici sono conservati in tre dimensioni (vedi anche Fig. 158) alla base della coda di Scipionyx samniti-
cus. Scala metrica = 5 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 155

Pelvic and hindlimb muscles

Even more remarkable than that of the intestine is the
preservation of somatic muscle bundles at the base of the
tail. In less than 20 mm?, that area contains the best pre-
served muscle tissue of Scipionyx: in fact, these bundles
are still arranged in compact fascicula composed of paral-
lel, strictly appressed cells fossilised in three dimensions
(Fig. 153). In addition, the distribution of muscle remains
and their association to skeletal elements in this area sug-
gests that, in topographical and functional terms, they rep-
resent at least three different pelvic and hindlimb muscles
(Figs. 117,153).

Puboischiofemoral muscle - The most cranial re-
mains of the pelvic musculature lie in strict contact with
the craniomedial face of the right ischial shaft (Fig.
154A). They consist of a residual patch shaped into a
rough drop by mechanical preparation, $ mm long and
2 mm wide, that at high magnification shows a whitish,
amorphous matrix, probably made of muscle-related
connective tissue in which several myofibres are embed-
ded (Fig. 154B). Apart from the preservation of well-
visible muscle cells, the most interesting feature is their
arrangement: the myofibres intimately contacting the
shaft of the ischium parallel the bone in a craniodorsal
direction; the distalmost ones begin almost parallel, then
curve in a cranial direction, giving the bundle a fan-like
appearance. There is no muscle scar or any other sign
of muscular insertion on the shaft of the ischium, but
proximal to the bone all myofibres converge towards the
cranial tip of the ischial foot.

Based on observations in a number of fossil taxa, and
using the extant phylogenetic bracket, Carrano & Hutch-
inson (2002) offered well-supported inferences concern-
ing most of the hindlimb musculature in 7yrannosaurus

rex. In Scipionyx, the bone surface in strict contact with
the muscle remains is topographically equivalent to the
one that in 7yrannosaurus would have given attachment
to the M. adductor femoris I (Carrano & Hutchinson,
2002). According to these authors, this muscle probably
originated from the cranioventral surface of the ischium,
then passed laterally and cranioventrally to insert on the
caudal surface of the femoral shaft, approximately two-
thirds of the way towards its distal end, where an oval,
rugose medial attachment site is located. In Scipionyx,
the cranial and caudal directions of the preserved my-
ofibres, and their fan-like arrangement that matches well
the anatomy of the M. adductor femoris I seen in extant
archosaurs (Hutchinson & Gatesy, 2000: fig 3A - errata
corrige: ADD2 and ADDI labels are inverted), are both
consistent with the two bone attachments suggested by
Carrano & Hutchinson (2002).

This adductor muscle is reconstructed by Carrano &
Hutchinson (2002) with a level I° inference (sensu Wit-
mer, 1995, 1997): it is present in Crocodylia and in Neor-
nithes, but the insertion on the cranioventral edge of the
ischium is slightly different in the two taxa — in Croco-
dylia, it is called M. adductor femoris I, whereas in Neor-
nithes it is the M. puboischiofemoralis pars medialis. As
for Tyrannosaurus, Carrano & Hutchinson (2002: fig. 3)
refer this muscle to the M. adductor femoris I, placing its
origin near the obturator process and leaving the ischial
shaft empty. Given the vicinity of the myofibres to the
bone and the lack of tendineous remains, Scipionyx sup-
ports the supposition (Carrano & Hutchinson, 2002) that
this muscle has a fleshy attachment. On the other hand,
given the convergent caudal orientation of the fibres, we
infer that the M. adductor femoris I does not originate on
or near the obturator process, but rather from the ischial
foot and a cranioproximal adjacent area. This evidence
also contradicts Paul (1988), who argued on functional

Fig. 154 - Scipionyx samniticus. A) remains of the puboischiofemoral musculature preserved on the shafts of the ischia; B) close-up
of a patch of muscle adjacent to the right ischial foot. Scale bars = 2 mm (A) and 200 um (B). See Appendix 1 or cover flaps for
abbreviations.
Fig. 154 - Scipionyx samniticus. A) resti della musculatura puboischiofemorale conservata lungo le ossa ischiatiche; B) particolare di
un lembo di muscolo adiacente il piede dell’ischio destro. Scale metriche = 2 mm (A) e 200 um (B). Vedi Appendice 1 o risvolti di
copertina per le abbreviazioni.

156 CRISTIANO DAL SASSO & SIMONE MAGANUCO

grounds that in theropod dinosaurs it was unlikely that
any muscle originated from the ischial shafî.

In conclusion, Scipionyx shows unequivocally that at
least one muscle inserts onto the ischial shaft of thero-
pods: this insertion is distal to the obturator process and,
more precisely, on a craniodistal area that, judging from
relatively recent descriptions and drawings (Carrano &
Hutchinson, 2002: fig. 3), is consistent with the M. pubo-
ischiofemoralis pars medialis of Gallus and the M. ad-
ductor femoris I of Tyrannosaurus. It cannot be excluded
that this muscle in 7° rex or in other theropods would have
had a craniodistal insertion that is more extended than the
one suggested in the literature.

Caudofemoral muscles - Evidence for the caudofem-
oral musculature in extinct archosaurs rests on skeletal
features associated with the origin and insertion of these
muscles (Gatesy, 1990) — except in Scipionyx. In fact,
three evident muscle bundles can be clearly seen with
the naked eye in the Italian compsognathid to parallel the
caudal vertebral column and reach cranially towards the
femora (Fig. 153). The most cranial ones consist of two
short bundles that emerge from the left side of the rectum
and run for about 5 mm below the 5° sacral centrum, dor-
sal and ventral to the faecal pellet. In a caudal direction,
the two bundles would seem to terminate abruptly at the
same level, but, in fact, the dorsal one continues under the
first caudal centrum, whereas the ventral one has an irreg-
ular margin clearly produced by mechanical preparation.
As outlined above, we exclude that these two bundles be-
long to rectal or cloacal muscles, as the craniocaudal di-
rection of the well-preserved myofibres is not consistent
with the circular arrangement of the sphincters. Proximity
to remains of the terminal tract of the rectum is not evi-
dence of any morphofunctional relationship with it, as is
topographically obvious for tail muscles: for instance, in
Alligator hatchlings (Dal Sasso & Maganuco, pers. obs.,
2009) the proximocaudal hypaxial musculature runs both
dorsal and ventral to the rectum and also very close to it. A
third muscle bundle, with the same direction and a similar
diameter (2 mm) but more distally placed, emerges from
the left side of the faecal pellet, i.e., from an innermost
plane, which underlies also the former bundles. This third
bundle is the best preserved in length, running in a caudal
direction for 15 mm, below caudal centra 1-3. Actually, its
caudal third is overlaid by another muscular and connec-
tive tissue mass (see below).

Given the common parallel fibre arrangement and di-
rection and the absence of myosepta, and given the deep,
left lateral position of all three bundles with respect to the
midline elements (proximal rectum, caudal vertebrae), we
refer these muscle remains to the left M. caudofemoralis.
M. caudofemoralis longus (CFL) represents the primary
femoral retractor muscle of lepidosaurs and crocodilians
(Gatesy, 1990); its smaller counterpart, M. caudofemoralis
brevis (CFB), is also likely important in hip extension. In
extant reptiles, the M. caudofemoralis is morphologically
unique among the caudal muscles in being not partitioned
by conical myosepta, its overall form more closely resem-
bling a limb muscle rather than an axial muscle (Persons
& Currie, 2011).

We encounter big problems when considering a pos-
sible attribution of the three muscle bundles of Scipionyx
to the CFB. In fact, such an attribution is incompatible

with the position of the preserved portions of the bundles
because even the two most proximal ones lie at some dis-
tance from the ilium, caudal to it, and display a fibre orien-
tation parallel to the long axis of the tail. As demonstrated
by Carrano & Hutchinson (2002) for 7yrannosaurus, the
CFB originated within the brevis fossa of the ilium, along
its ventral edge caudal to the acetabulum, and extended
cranioventrally to insert on the caudolateral shaft of the
femur.

As in extant crocodilians and lepidosaurs, the CFL in
Tvrannosaurus and in Scipionyx was probably the largest
single muscle of the hindlimb. According to Carrano &
Hutchinson (2002), this muscle originated from the lat-
eral faces of the proximal caudal vertebral centra, pass-
ing cranioventrally to insert on the medial portion of the
fourth trochanter. Proximity to caudal centra and cranial
orientation of the myofibres in Scipionyx are consistent
with this arrangement, i.e., with a muscle that would have
filled the ventrolateral sulcus created by the transverse
processes and the chevrons (Gatesy, 1990).

The three-dimensional fossilisation of the M. caudo-
femoralis is apparent to the naked eye, but it becomes
dramatic at high magnification. The preservation is so
exquisite that it allows not only to distinguish the indi-
vidual myofibres under the optic microscope (Fig. 155),
but also subcellular details with scanning electron miero-
scopy. Under the optic microscope, the myofibres have
an intensely brown-reddish colour, and the intercellular
spaces are diaphanous; if brightly illuminated, the crystal-
lised tissue allows some vision through it (Fig. 156). Like
the myofibres preserved at the base of the neck, those of
the caudofemoral musculature are parallel, straight and
stacked in multiple layers, as in most somatic muscles of
extant vertebrates. Therefore, although they vary in size
from one another in the different portions of the three
muscle bundles (10-26 um), the diameter of each indi-
vidual myofibre is very constant.

Fig 155 - Close-up taken under the optical microscope of the caudo-
femoral myofibres, still forming a compact muscle bundle just ventral
to the 1% caudal vertebral centrum of Scipionyx samniticus. Scale bar
= 200 um.

Fig 155 - Scipionyx samniticus. Particolare al microscopio ottico delle
miofibre caudofemorali, che sono ancora unite a formare un fascio
muscolare compatto, sotto il 1° centro vertebrale caudale. Scala me-
trica= 200 um.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY ST7

Fig. 156 - The cellular-level preservation in the caudofemoral muscles of Scipionyx samniticus can be seen even with a camera
equipped with a macro lens. Scale bar = 200 um. See Appendix 1 or cover flaps for abbreviations.

Fig. 156 - Nei muscoli caudofemorali di Scipionyx samniticus la conservazione dei tessuti molli a livello cellulare è visibile anche con
una fotocamera con obiettivo macro. Scala metrica = 200 um. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

SEM imaging reveals the spectacular preservation of
a sarcomere-related banded pattern within every single
mvyofibre (Fig. 157A-C). The myofibres are polygonal in
transverse section, with individual margins that are well-
delimited by narrow spaces left by the dissolution of the
cell membrane (sarcolemma) and its connective tissue
sealing (endomysium); the striations of the sarcomeres are
continuous, regular, complete and evident, as seen only in
muscle tissue from some invertebrates and fishes from the
Upper Jurassic of Cerin (France) and Solnhofen (Germa-
ny) (Briggs, 2003: fig. 7a). The sarcomere periodicity (2.5
um) is consistent with the mean sarcomeric size in extant
vertebrates (e.g., Muhl et al., 1976; Cross et al., 1981;
Botte & Pelagalli, 1982; Wheeler & Koohmaraie, 1994;
Cavitt et al., 2004; Panchangam et al., 2008).

At the highest magnification that can be used to see
the myofibres of Scipionyx, the myofibrils can be seen to
have been replaced by characteristic ‘“furry” microspheres
and coated by dispersed euhedral apatite crystals (Fig.
157C), just like in some samples from decapod crusta-
ceans published by Wilby & Briggs (1997: figs. 3d, 4d).
In the literature, the only other dinosaurian muscle tissue
with a similar appearance is found in Santanaraptor from
the Lower Cretaceous Santana Formation of Brazil (Kell-
ner, 1996a,1996b); however, different to Scipionyx, it has
incomplete sarcomere-related bands, described by the au-
thors as “partial striations” (Kellner, 1996a). Moreover,
considering that the average diameter of the myofibres in
the Brazilian specimen is about four times that of Scipio-
nyx 5, preservation at ultrastructural level is definitely
better in the Italian compsognathid.Three-dimensional
myofibres, preserving polygonal cross-sections, trans-
verse striations and even perimysial or epimysial remains
are also described by Chin ef a/. (2003) in the fossilised
muscle tissue from an indeterminate vertebrate, possibly
a pachycephalosaur, embedded within a possible tyranno-
saurid coprolite from the Late Cretaceous of Canada.

AIl SEM samples from the caudofemoral musculature
of Scipionyx show the same spectacular level of preserva-
tion of the sarcomeres, and their banded pattern can be
recognised even in non-conventional views: for exam-

ple, in a relatively peripheral sample from the dorsalmost
bundle (Fig. 157E), the ultrastructural aspect of the sar-
comeres differs in appearance only because of its oblique,
rather than longitudinal, cutting plane. As a matter of fact,
our sample looks strikingly similar to obliquely sectioned,
three-dimensionally preserved fossil muscles illustrated
by Wilby & Briggs (1997: figs. 1d-e).

The exceptionally detailed preservation of the caud-
ofemoral muscle tissue of Scipionyx, as well as the SEM
element microanalysis of this tissue (Fig. 157D), indi-
cate that, after the hatchling theropod died, its carcass
was subjected to very little decay and to rapid minerali-
sation in the presence of a high concentration of phos-
phates; diagenetic processes were halted very early on
and fossilisation outpaced degradation through a process
called authigenic mineralisation (Briggs, 2003), which
will be discussed in the section devoted to soft tissue
taphonomy.

Ilio-ischiocaudal muscle septa - Several hollow
filamentous traces, embedded in connective tissue and
seemingly related to muscle remains, lie under the tail of
Scipionyx (Figs. 153, 158, 159). Despite this, these trac-
es differ definitely in aspect, structure and size from the
surrounding muscle fibres, as well as from those of the
neck. In fact, these hollow structures are dark coloured
and taper to a pointed end both proximally and distally,
but in a distal direction they exhibit a bifurcated, V-
shaped, often imbricate pattern; they show some flexural
stiffness for most of their length, but have wavy distal
ends (Figs. 159A-B; 160A-B). Moreover, the ventral-
most structures become shorter and thinner: their length
ranges from 9 mm (dorsal structures) to 4-5 mm (ventral
structures), and their diameter ranges from 0.16 mm to
0.06 mm, respectively.

Several interpretations have been tentatively hypothe-
sised for these enigmatic structures. Dal Sasso & Signore
(1998a), observing a seeming connection with the caudo-
femoral bundles (Fig. 161 — see in particular the proximal-
most point in which the filaments overlap onto the vent-
rolateral CFL bundle; see also below), wrote about large

1 58 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Tazsore=29 945 ke Gent ID=
Vert= 1000 Window 0.005 - d1955= 62494 cn

100H:m

Fig. 157 - SEM imaging of the dorsalmost bundle of the caudofemoral muscle of Scipionyx samniticus: A) fragment of a fasciculum
of myofibres; B) close-up of A; C) close-up of an isolated myofibre. While sampling the specimen, longitudinal partition occurred
along natural intercellular spaces, like în vivo, except for one chip-like fracture (A, centre-right). D) SEM element microanalysis of the
myofibre shown in C. E) the sarcomere-related banded pattern is observable also when the myofibres are cut obliquely. In taphonomy,
such a level of preservation is related to the formation of a substrate microfabric (sensu Wilby & Briggs, 1997). See Appendix 1 or
cover flaps for abbreviations.

Fig. 157 - Immagine al SEM del fascio più dorsale di muscolo caudofemorale di Scipionyx samniticus: A) frammento di un fascicolo
di miofibre; B) particolare di A; C) particolare di una miofibra isolata. Nel prelievo del campione, il distacco in senso longitudinale è
avvenuto lungo gli spazi intercellulari naturali, come in vivo, fatta eccezione per una frattura di aspetto scheggiato (A, centro-destra).
D) microanalisi degli elementi al SEM della miofibra mostrata in C. E) il taglio obliquo delle miofibre non nasconde la struttura a
bande derivante dalla fossilizzazione dei sarcomeri. In tafonomia, un tale livello di conservazione identifica una “substrate micro-
fabric“ (sensu Wilby & Briggs, 1997). Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

muscle fibres (Dal Sasso & Signore, 1998a: fig. 2). Given
the different aspect, size and kind of preservation of all
other muscle fibres (more properly, myofibres) preserved
in the specimen and described in detail herein, today this
interpretation must be rejected.

Tendons were another possible anatomical interpreta-
tion. A lattice of ossified tendons is found in the hypaxial
region of ornithopod dinosaur tails (e.g., Organ, 2006).
However, in this case one would expect these structures
to lie mediodorsally to the muscle bundles, because ten-
dons attach directly to the bones, whereas the structures

of Scipionyx lie in a position usually occupied by a tail
muscle, the M. ilio-ischiocaudalis (Persons & Currie,
2011). The non-bifurcating pattern of ornithopod tendons
is another important character that does not fit with these
enigmatic structures of Scipionyx.

A third hypothesis was made by Currie (pers. obs.,
1999), who observed that the filaments of Scipionyx
look like the protofeathers of some Chinese compsogna-
thids (e.g., Zhou et al., 2003; Xu, 2006). The position of
the filaments is consistent with this interpretation: they
might belong to an integumentary external covering be-

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Fig. 158 - A grazing view of the sofî tissues fossilised ventral to the tail of Scipionyx samniticus highlights the hollow structures pos-
sibly related to ileo-ischiocaudal muscle septa. Note also the three-dimensional preservation of the caudofemoral muscle bundles. See
Appendix 1 or cover flaps for abbreviations.

Fig. 158 - Scipionyx samniticus. Una vista radente dei tessuti molli fossilizzati sotto la coda evidenzia le formazioni cave probabil-
mente legate ai setti del muscolo ileo-ischiocaudale. Si noti anche la conservazione tridimensionale dei fasci muscolari caudofemorali.
Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

Fig. 159 - The ?ileo-ischiocaudal muscle septa of Scipionyx samniticus, seen under ultraviolet-induced fluorescence.
Fig. 159 - I ?setti muscolari ileo-ischiocaudali di Scipionyx samniticus visti in fluorescenza indotta da luce ultravioletta.

160 CRISTIANO DAL SASSO & SIMONE MAGANUCO

agitano Ti aL Lot

Fig. 160 - A) ?ileo-ischiocaudal muscle septa of Scipionyx samniticus under visible light. B) close-up of some of the hollow, axially-
bifurcated traces of the possible septa shown in A, embedded in connective tissue and muscle remains. C) close-up of the septa of the
M. ilio-ischiocaudalis of a juvenile Caiman crocodilus (the arrow points to cranial direction). Scale bars = 2 mm (A), 0.5 mm (B) and

5 mm (C). See Appendix 1 or cover flaps for abbreviations.

Fig. 160 - A) ?setti muscolari ileo-ischiocaudali di Scipionyx samniticus fotografati in luce visibile. B) particolare di alcune tracce cave
dei probabili setti mostrati in A, che appaiono biforcate lungo l’asse principale e immerse in tessuto connettivo e resti di muscolatura.
C) particolare dei setti del M. ilio-ischiocaudalis di un giovane Caiman crocodilus (la freccia è rivolta in direzione craniale). Scale
metriche = 2 mm (A), 0,5 mm (B) e 5 mm (C). Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

cause they lie some distance from the vertebrae. At first
glance, this hypothesis seems supported by superficial
similarities, such as that the filaments of Sinornithosau-
rus (Xu et al., 1999) are hollow tubes and that the ones
of Dilong (Xu et al., 2004) exhibit a roughly comparable
branching pattern. Nevertheless, at closer examination
also the protofeather hypothesis must be abandoned. In
fact, the enigmatic structures of Scipionyx gently taper
not only in a distal, but also in a proximal, direction, a
fact that would be a structural and functional nonsense
for any integumentary derivative: the proximal end is
for dermal attachment, so needs to be a firm, robust, en-
larged base (calamus, hair bulb). According to Xu (pers.
comm., 2009), the counterpart-like preservation, as well
as the axial bifurcation and irregular tapering shown by
the hollow structures of Scipionyx, are not consistent
with the protofeathers of the Chinese dinosaurs, also be-
cause protofeathers are not comparable dimensionally,
being thinner and more delicate to the point that they
cannot produce any impression in the sediment. More-
over, they branch repeatedly and always from a single
central axis.

A fourth hypothesis arose during the present study,
after we examined the hollow structures of the abdomen
that we then referred to blood vessels of the cranial me-
senteric artery. However, the proximal and distal tapering
we observe in these enigmatic structures does not make
any sense as part of a vessel either, given that a vessel
is made to drive a liquid through. Comparison with the
blood vessels of the cranial mesenteric artery strengthens
our rejection of this hypothesis because, in addition, the
hollow structures under the tail are longer, straight rather
than wavy, much more strongly coloured and not at all
diaphanous. Furthermore, most of them have a greater di-
ameter than the intestinal blood vessels.

The fifth and, in our opinion, most feasible hypothesis,
came from an exchange of information that began recently
(Persons, pers. comm., 2009, 2010). In a variety of extant
reptiles, M. caudofemoralis is covered by M. ilio-ischio-
caudalis, which wraps around the former laterally and
ventrally for all its length (e.g., Huchzermeyer, 2003; Per-
sons & Currie, 2011). Consistent with its nomenclature, M.
ilio-ischiocaudalis is composed of two major muscles: M.
iliocaudalis and M. ischiocaudalis, with the former origi-

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 161

nating from the ilium and the latter from the ischium. M.
ilio-ischiocaudalis is relatively thin proximally, whereas
distally it increases in relative thickness as the thickness of
M. caudofemoralis decreases. After the latter disappears,
M. ilio-ischiocaudalis inserts on the full lateral surface of
the caudal centra and the haemal arches. Contrary to the
caudofemoral muscle, which consists of straight paral-
lel bundles, the ilio-ischiocaudal muscle is partitioned by
conical myosepta, that on the surface, or in a longitudi-
nal cross section, produce a pattern of craniocaudally V-
oriented bundles. These septa bear a strong resemblance
(Fig. 160C; Persons, pers. comm., 2009) to the imbricate
V-shaped structures of the enigmatic structures of Scipio-
nyx running parallel to the caudofemoral muscle bundles
and lying laterally and ventrally, but not medially, to them.
The dimensional range of these structures in individuals of
extant reptiles of comparabile size is also consistent. There-
fore, we conclude that the hollow enigmatic structures of
Scipionyx are most likely ilio-ischiocaudal septa. Specifi-
cally, the well-defined septate pattern on the dorsalmost
portion might belong to the right M. iliocaudalis, whereas
the ventral portion, where the pattern is less well-defined,
would belong to the right M. ischiocaudalis.

In this perspective, the “glued” appearance of the
?right ilio-ischiocaudal muscle on the ?left caudofemo-
ral muscle (Fig. 161) might be due to osmosis-driven de-
hydration that possibly affected the carcass of Scipionyx
when it sunk in the Pietraroja basin.

Fig. 161 - Close-up ofthe contact that seems to exist between the caudal-
most bundle of the ?left caudofemoral muscle and the tapering cranial
end of the cranialmost ?septum of the ?right ilio-ischiocaudal muscle.
Scale bar = 1 mm. See Appendix 1 or cover flaps for abbreviations.
Fig. 161 - Contatto apparente tra il fascio più caudale del muscolo cau-
dofemorale ?sinistro e la terminazione affusolata del ?setto muscolare
ilio-ischiocaudale ?destro più craniale. Scala metrica = 1 mm. Vedi
Appendice 1 o risvolti di copertina per le abbreviazioni.

Other fragments of pelvic and hindlimb muscles -
At least six relatively loose muscle fragments, composed
of parallel myofibres of comparable aspect and size, are
preserved in the pelvic and proximal caudal region of
Scipionyx (Figs. 117, 153). The three cranialmost frag-
ments are superimposed, respectively, on the descending
tract of the rectum and on the ascending tract nearby (e.g.,
Fig. 162). Their distance from the vertebral column as well
as the craniocaudal direction of their myofibres are con-
sistent with the position along which the right M. caudo-

Fig. 162 - Close-up of myofibres covering the rectum of Scipionyx sam-
niticus, between the 5° sacral centrum and the ischial feet, tentatively
referred to the right caudofemoral muscle. Scale bar = 200 um. See
Appendix 1 or cover flaps for abbreviations.

Fig. 162 - Scipionyx samniticus. Particolare di un frammento di tessuto
con miofibre, appoggiato sul retto tra il 5° centro vertebrale sacrale e
i piedi ischiatici, attribuito tentativamente al muscolo caudofemorale
destro. Scala metrica = 200 um. Vedi Appendice 1 o risvolti di coper-
tina per le abbreviazioni.

femoralis or the right M. ischiocaudalis would have run.
The former is preferred, as the three fragments lie directly
on the surface of the rectum, i.e., in a more medial plane
than M. ischiocaudalis (e.g., Gatesy, 1990: fig. 3a; Per-
sons & Currie, 2011: fig. 2a). Attribution to M. adductor
femoris I is unlikely: as already pointed out (Carrano &
Hutchinson, 2002), this muscle originates on the caudal
edge of the near ischium, but it runs cranioventrally, not
caudally, to insert on the femur. M. ischiotrochantericus is
another possible candidate, but this is improbable because
it originates from the medial surface of the ischium, which
is only a few millimetres cranial to these fragments. In
fact, in crocodilians and other extant reptiles, this muscle
also inserts onto the caudolateral surface of the proximal
femur, then runs craniodorsally to the ischium, rather than
caudally (Carrano & Hutchinson, 2002).

Three other muscle fragments are embedded in the
connective tissue including the supposed ilio-ischiocau-
dal septa (e.g., Fig. 163). AII three fragments are found
at the level of the 3'° caudal centrum, subequally distant
to each other. The most dorsal one seems in line with the
left M. caudofemoralis, so might represent its caudal con-
tinuation; the ventralmost one, and the one in-between the
other two, are too ventrally placed to be included in the
caudofemoral muscle; thus, they might be remnants of the
right ischiocaudal myofibres, for the same reasons we in-
voked when regarding as possible ischiocaudal septa the
nearby mysterious structures.

Caudal hypaxial connective tissue/ligaments

In between the two better exposed chevron bones,
which are preserved in an intercentral position between
caudal vertebrae 4-5 and 5-6, a small patch of soft tissue
can be detected under visible light at medium magnifica-
tion, and fluorescing under UV light (Fig. 164A). High

| 62 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 163 - Close-up of myofibres preserved in between the hollow fila-
mentous traces illustrated in Figs. 159-160, possibly belonging to the
right ischiocaudal muscle of Scipionyx samniticus. Scale bar = 200 um.
See Appendix 1 or cover flaps for abbreviations.

Fig. 163 - Particolare di un frammento di tessuto con miofibre con-
servato fra le tracce filamentose cave illustrate nelle Fig. 159-160, che
potrebbe appartenere al muscolo ischiocaudale destro di Scipionyx
samniticus. Scala metrica = 200 um. Vedi Appendice 1 o risvolti di
copertina per le abbreviazioni.

magnification (Fig. 164B) reveals a number of light brown
fibres, possibly myofibres, and a couple of brown coloured,
flat, arched structures in the centre of the area, paralleling
the curvature of the chevron shafts. These myofibres and
arched structures are embedded in a yellowish, amorphous
material. The flat, arched structures have a transverse di-
ameter (0.2 mm) which is evidently larger than that of any
mvyofibre preserved here or in other parts of the specimen,
and a fibro-laminous aspect which is not even consistent
with a bundle of tightly bound myofibres. Rather, such an
aspect is compatible with a sheet-like single unit.

According to Persons (pers. comm., 2009), a tendon
latticework is expected to run in dinosaurs from haemal
arch to haemal arch, medial to the tail muscles but just
lateral to the chevron bones, similar to what is seen in ex-
tant reptiles. However, the laminae preserved in Scipionyx
seem to lie just in the medial sagittal plane, and their thin
delicate aspect is not at all reminiscent of tendons. So we
are confident in referring them to the fibrous connective
tissue sheets that separated the counterlateral caudofemo-
ral muscle bundles along the midline, or more likely to the
ligamentum interhaemale (Frey, 1988; Schwarz- Wings et
al., 2009). The amorphous, whitish material surrounding
the laminae might belong to loose connective tissue as-
sociated with them.

Indeterminate ?connective tissue

Some patches of soft tissue preserved in Scipionyx are
apparent as a yellowish material embedding myofibres,
myosepta and/or fibrous connective tissue. We often de-
scribe this material as an amorphous mass, because it does
not have any particular structural arrangement under the

Fig. 164 - Remnants of the caudal hypaxial connective tissue, or ligaments, in Scipionyx samniticus, under ultraviolet-induced fluores-
cence (A) and visible light (B, close-up). Scale bar = 0.5 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 164 - Scipionyx samniticus. Residui del connettivo o dei legamenti ipassiali caudali, fotografati in fluorescenza indotta da luce ultra-
violetta (A) e in luce visibile (B, particolare). Scala metrica = 0,5 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 163

optical microscope (e.g., Figs. 129, 159A). The largest
areas where this tissue is preserved are found at the base
of the neck and at the base of tail — shown in grey-brown
shading in the soft tissue general map (Fig. 117).

The possibility that this material represents diageneti-
cally altered adipose tissue cannot be ruled out, because
saturated fatty acids are resistant to degradation just like
connective tissue (Chin et a/., 2003; Killops & Killops,
1993). Nevertheless, fat storage in a hatchling is unlikely,
so we have reason to suppose that these tissue patches (es-
pecially the one at the base of the tail) have a connective-
tissue nature, based on their co-association and intimate
connection with myofibres, myosepta and fibrous connec-

tive tissue. Unfortunately, samples of this material were
not collected, so at present we cannot test for the presence
of the collagen-diagnostic 67 nm banding pattern, given
by the peculiar overlap of the molecules composing the
collagen fibrils, as conducted by Schweitzer et a/. (2007,
2008) on dinosaurian soft tissue. Certainly, nothing simi-
lar can be seen under the light microscope, and the aspect
of this material remains amorphous. This might be due
not only to instrumental limits but also to partial decay
of the tissue before fossilisation, so that we cannot dis-
tinguish whether it was arranged as fibrous or loose con-
nective tissue in vivo. For all these reasons, we label this
tissue as indeterminate.

EXTERNAL SOFT TISSUES

This section describes integumentary remains, more
precisely the horny structures (claws), that from a histo-
logical point of view are not a kind of tissue but epider-
mal derivatives originally composed of keratin. Although
keratin is a highly resistant insoluble protein, fossilisation
usually takes place after the loss of the skin and its deriva-
tives, so that complete horny claws in fossil vertebrates
are found very rarely. Among theropod dinosaurs, the
first recording was in Archaeopteryx (Ostrom, 1970), but
apart from single finds, such as Rahonavis (Schweitzer
et al., 1999) and Santanaraptor (Kellner, 1999), most of
the other examples came in recent times from the excep-
tionally preserved integuments of the “feathered” Yixian
dinosaurs (e.g., Microraptor, Protarchaeopteryx, Sinorni-
thosaurus).

Horny claws

AIl preserved manual ungual phalanges of Scipio-
nyx are sheathed by scythe-like, pointed horny claws,

greatly prolonging the curvature of the bony elements
(Fig. 165). The horny claws are still well-attached to
their bony phalanges. The base attachment is so firm that
the horny claws did not detach, even where the pointed
apexes bent partially (digits II and III of the left manus)
or fully (digit I) towards the ventral margin of the bony
ungual phalanges (Fig. 166). According to Delfino (pers.
comm., 2006), this is evidence of rapid burial, because
when the carcasses of extant reptiles are left soaking, the
horny claws become quickly detached, together with the
skin of the limbs. Therefore, we can deduce that, after
Scipionyx died, it did not float for any length of time.
The horny tip of the 3" right digit seems to be in
its original apical position by comparison with the other
ones. What looks strange is the large horny area embrac-
ing it, the size of which seems to fit a base attachment
rather than a distal portion of a pointed claw. Actually,
this large horny area represents the dorsal and ventral
walls of the talon that opened-up post mortem, thus ex-
posing their internal surface as an almost flattened table
(Fig. 167). As for dimensions, the horny claws of Scipio-

Fig. 165 - Horny claws are still in place on all preserved left (A) and right (B) manual unguals of the holotype of Scipionyx samniticus.
Scale bar = 5 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 165 - Gli artigli cornei sono ancora in posto su tutte le falangi ungueali conservate nella mano sinistra (A) e destra (B) dell’olotipo
di Scipionyx samniticus. Scala metrica = 5 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

164 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 166 - Close-up of the 1% left manual ungual of Scipionyx samniti-
cus, showing the pointed apex of the horny claw broken-off and bent-
back against the ventral margin of the bone. Scale bar = 0.5 mm. See
Appendix 1 or cover flaps for abbreviations.

Fig. 166 - Particolare della 1° falange ungueale della mano sinistra di
Scipionyx samniticus, che presenta la punta dell’artiglio corneo ripiegata
completamente contro il margine ventrale dell’osso. Scala metrica = 0,5
mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

Fig. 167 - The unusual state of preservation of the 2°‘ right horny claw
of Scipionyx samniticus. Scale bar = 1 mm. See Appendix 1 or cover
flaps for abbreviations.

Fig. 167 - L’inusuale modalità di conservazione del 2° artiglio corneo
della mano destra di Scipionyx samniticus. Scala metrica = 1 mm. Vedi
Appendice 1 o risvolti di copertina per le abbreviazioni.

nyx vary in length from 7.7 mm (linear length of the 2°‘
left claw) to 6.2 mm (linear length of the 3'° right claw),
out-measuring the bony manual phalanges up to 3.2 mm
(linear length of the 2°° left claw beyond its bony tip).
This means that, when reconstructing the in vivo aspect
of a theropod hand, the length and the curvature of the
fossil bony claws distal to the flexor tubercles should be
increased by some 40%.

The preservation of the horny claws of Scipionyx is
so good that under the light microscope a peculiar semi-
transparent, waxy aspect can be seen. No thin sectioning,
nor immunohistochemical or mass spectrometric studies
were possible, so we were unable to investigate the micro-
structure and composition of the horny claws of the Ital-
ian compsognathid, as has been done on other theropod
material (e.g., Schweitzer ef a/., 1999). Nevertheless, in
addition to its waxy and fibrous aspect, it is also evident
that the tightly adhering material of the bony claws is ar-
ranged in layers, and that the layers composing the dor-
sal horny talons (unguis) are darker in colour than those
composing the ventral horny talons (subunguis). Further-
more, the dark colour of the dorsal horny talons increases
towards the tips, becoming almost black (Fig. 168). This
feature indicates a higher density of material and, thus, a
proportionate mechanical resistance, as is the case in most
extant tetrapods in which the unguis is known to have a
hard, thick dorsal layer, whereas the subunguis has a soft,
thin ventral layer.

Remarkably, a similar colouring is present also in the
fossil Santanaraptor: like Scipionyx, the Brazilian thero-
pod has soft tissue that can be distinguished from the
bones by their lighter colours, except near the unguals,
where the horny covering is reported to be darker than the
bone (Kellner, 1999).

Fig. 168 - Close-up of the 1% right manual horny claw of Scipionyx
samniticus. Note that the colour of the unguis becomes darker towards
the pointed apex. Scale bar = 1 mm. See Appendix 1 or cover flaps for
abbreviations.

Fig. 168 - Particolare del 1° artiglio corneo della mano destra di Scipio-
nyx samniticus. Si noti la colorazione scura dell’unguis, che aumenta
in direzione della punta. Scala metrica = 1 mm. Vedi Appendice 1 0
risvolti di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 165

ENDOGENOUS BIOMOLECULAR COMPONENTS

Preservation of endogenous biomolecules has been
documented in a variety of vertebrate and invertebrate fos-
sil specimens from varying ages (Schweitzer ef a/., 1999).
Schweitzer ef al. (2008) demonstrated that, although the
almost perfect preservation of form does not guarantee
the presence of original organic material, the preserva-
tion of endogenous molecules is closely correlated to it.
The fossilisation of different types of tissues in different
colours also suggests different chemical composition (Sch-
weitzer, pers. comm.,1998); therefore, some endogenous
biomolecular components might still be locally present in
Scipionyx. For instance, highly resistant insoluble kera-
tin molecules might be among the chemical components
of the above-mentioned dark portion of the horny claws;
some connective tissue patches might contain collagen
remains; haem components derived from haemoglobin
molecules or their products of decay might be preserved in
the reddish pigments of the liver. According to the Univer-
sity of South Florida’s endocrinologist David Vesely (pers.
comm., 2008), the intestine of Scipionyx might even con-
tain atrio-natriuretic peptides, vasodilator hormones which

are usually released by the heart but are also concentrated
in the gastrointestinal tissues. Nevertheless, this specula-
tion remains such, as neither chemical analyses (other than
SEM element microanalysis) nor molecular analyses were
undertaken in this study because the majority of presently
used extraction techniques — destruction of macroscopic
samples through crushing to a fine powder, with multiple
repetitions of the process (Schweitzer ef a/., 2008) — would
cause serious damage to this unique specimen.

Biomolecules have a preservation potential that depends
on their size and complexity: for instance, in fossil specimens,
macromolecules are found to be more resistant to decay than
proteins or nucleic acids (Tegelaar et a/., 1989; Briggs, 1993;
Schweitzer et al., 2008). The presence»in Scipionyx of origi-
nal autogenous DNA should be, anyhow, excluded, given the
nature of the deposits from which the fossil comes. Natural
biological caskets, like the thick bony wall of a 7prannosau-
rus femur, have been demonstrated to be much better sites
of biomolecular preservation than the laminated limestones,
probably because the latter are subjected to intense water
permeation (Schweitzer et al., 2007).

Table 3 - Selected measurements of the soft tissues of the holotype of Scipionyx samniticus.

Muscles and other soft tissues in the neck
Neck muscles and other soft tissue
i 19.3 mm
remains, length
Neck muscle and other soft tissue
: 12.6 mm
remains, width
Neck muscle myofibres, 8
du i um
minimum diameter
Neck muscle myofibres,
13 um
maximum diameter
Neck muscle myofibres, 9:00
sarcomeric periodicity (length) SAS
Neck ?collagen bundles, 8
ve. i um
minimum diameter
DI
Neck ?collagen bundles, 40 um
maximum diameter
Trachea
Trachea, length of preserved tract 7 mm
Number of preserved tracheal rings 10-11
L
Tracheal rings, average diameter I
mm
of the preserved parts
Tracheal rings, average craniocaudal 03
length
Space between rings,
average craniocaudal length MANAIDT

Oesophagus
Oesophagus remains, length Smm
Liver, and other blood-rich organs
Reddish halo, diameter lan
(under UV light)
Intestine
Intestine, visible length 173 mm
Intestine, estimated length 300-320 mm
Duodenum, visible length 72mm
Duodenum, estimated length 105-115mm
Duodenum, maximum diameter 6.5 mm
Duodenum, average diameter 5.2 mm
Number of duodenal folds
i 5-6

(plicae circulares) per mm
Longitudinal visceral (duodenal) 7

ir i um
myofibres, minimum diameter
Longitudinal visceral (duodenal)

17 um

myofibres, maximum diameter
Jejunum, visible length 57 mm
Jejunum, estimated length 85 mm

166

CRISTIANO DAL SASSO & SIMONE MAGANUCO

Jejunum, maximum diameter

Jejunum, average diameter

Number of jejunal folds
(plicae circulares) per mm

Rectum, visible length

Rectum, estimated length

Rectum, maximum diameter

Rectum, average diameter

Faecal pellet, visible length

Faecal pellet, estimated length

Faecal pellet, maximum diameter

Faecal pellet, average diameter

Mesenteric blood vessels

Largest blood vessel, length

Largest blood vessel,
minimum diameter

Largest blood vessel,
maximum diameter

Puboischiofemoral muscle

Muscle remains, length

Muscle remains, diameter

Myofibres, minimum diameter

Myofibres, maximum diameter

Caudofemoral muscle (M. caudofemoralis longus)

Dorsalmost bundle, visible length

Dorsalmost bundle, diameter

Dorsalmost bundle myofibres,
| minimum diameter

Dorsalmost bundle myofibres,
maximum diameter

Dorsalmost bundle myofibres,
sarcomeric periodicity (length)

Ventralmost bundle, visible length

Ventralmost bundle, diameter

Smm Ventralmost bundle myofibres,
fo 10 um
minimum diameter
3.5 mm
Ventralmost bundle myofibres,
. 22 um
6 maximum diameter
Distalmost bundle visible, length 9.8 mm
35.5 mm i
Distalmost bundle, diameter 2.2 mm
55-65 mm
Distalmost bundle myofibres, 1000
Smm minimum diameter H
1
4 mm Distalmost bundle myofibres, DE
i um
maximum diameter
5.5 mm
6-7 mm
?Ilio-ischiocaudal muscle ?septa
2.5 mm J
IRA Length of preserved area 17 mm
Width of preserved area 11.5 mm
Smallest element, length 4-5 mm
OTO STA Largest element, length 9 mm |
Smallest element, maximum diameter 0.06 mm
0.04 mm
Largest element, maximum diameter 0.16 mm
0.1 mm
Horny claws Left Right
Manual digit I horny claw, immi 72700
linear length
S mm ì
Manual digit I horny claw,
2 mm oversize 3.2 mm| 2.8.mm
(OA (linear length beyond bony tip) |
Manual digit I horny claw,
15 um height at bony tip 0.7 mm | 09mm
|
Manual digit II horny claw, TAI È
linear length
6.3 mm Manual digit Il horny claw,
oversize 3.3 mm -
1.6 mm (linear length beyond bony tip)
Le T I
Manual digit II horny claw,
13 lm height at bony tip LIT i
Manual digit III horny claw,
26 um linear length 6.7 mm | 6.2 mm
Soi Manual digit III horny claw,
oVersize 2 mm -
(linear length beyond bony tip)
4.3 mm
Manual digit III horny claw, 08m
22 mm height at bony tip pia A

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 167

SOFT TISSUE TAPHONOMY

Exceptional preservation of labile soft tissue, such
as muscles or loosely connected integumentary struc-
tures (see Horny Claws) indicates that diagenetic
processes were halted very early after Scipionyx died
and that mineralisation outpaced degradation (Sch-
weitzer et al., 2008). Briggs (2003) described a pe-
culiar process, that he named authigenic mineralisa-
tion, in which microscopic mineral crystals replicate
the morphology of a biological structure, driven by
the activity of decay bacteria. This process represents
the highest taphonomic threshold among the possible
stages of soft tissue preservation. According to Briggs
(2003), the best known examples are the fish muscle
tissue from the Lower Cretaceous Santana Formation
of Brazil, which preserve sarcolemma and sarcomere-
related banding (Martill, 1988; Briggs ef a/., 1993).
Such details of labile soft tissue may survive as a re-
sult of the precipitation of authigenic minerals on the
tissues themselves, or on microbial films that cover
them. The relevant role of bacterial sealing in the ini-
tial stages of fossilisation was confirmed by experi-
mental simulations (Sagemann et a/., 1999; Iniesto
et al., 2009; Noto, 2009). The quality and fidelity of
preservation depends on the composition, form and
size of the mineral crystals. Authigenic mineralisation
also involves the invasion of tissues by ion-bearing
(e.g., phosphorus) pore water.

Scipionyx was preserved in a marine environment.
If the model of the shallow lagoon is correct (see
Geological Setting), some additional factors, such as
reduced vertical water stirring and seasonal overheat-
ing, might have favoured low oxygen levels in the
Pietraroja deposits (Bravi & Garassino, 1998), slow-
ing down the decay of the dinosaur carcass. Nonethe-
less, we would like to remark that in aquatic settings,
even under normal conditions (i.e., open waters), the
oxic-anoxic boundary may be at, or even above, the
sediment-water interface, and that underwater, most
decay of carcasses is usually anaerobic (Briggs, 2003).
Recent experiments confirm this statement and our hy-
potheses on soft tissue preservation in Scipionyx (see
below). Noto (2009), for instance, performed a year-
long taphonomic experiment investigating the effect of
different environmental conditions on the diagenesis of
buried bone, observing authigenic mineral formation.
Bone was also found to interact with the surrounding
sediment, buffering pore water pH changes and con-
tributing to local anoxia.

The major process of organic matter oxidation in
marine sediments is bacterial sulphate reduction (Gail-
lard et al., 1989), which can result in the precipitation
of minerals such as calcite, aragonite and apatite. Early
cementation is essential for the preservation of soft tis-
sue in three dimensions like we see in Scipionyx. In
fact, rapid authigenic mineralisation via phosphati-
sation can occur virtually immediately after death in
an appropriate environment (Briggs & Kear, 1993b),
preserving cells and subcellular detail faster than they
could decay (Butterfield, 2002). Authigenic mineralisa-
tion occurs through early infiltration and permeation of
labile tissue by mineral-charged water and differs from
petrification, which is a replacement process.

In a given specimen, however, these processes do
not take place in the same way, at the same time or with
the same kinetics. In fact, as well-remarked by Briggs
(2003), carcasses are made up of a range of materials
with different degrees of resistance to decay, and even
at the molecular level, different organic materials have
different potential for preservation. Moreover, the pre-
cipitation of one mineral or another is linked to labile
switches in the chemical reactions. For example, pH is
a major determinant in the control of calcium carbonate
and calcium phosphate precipitation, operating at the
microbial level as an “on” or “off” switch in the differ-
ent parts of the same carcass (Briggs & Wilby, 1996;
Noto, 2009; Sagemann et al., 1999). In our opinion,
this is the most likely explanation for preservation of
the intestine and decay of the stomach in Scipionyx.
As we pointed out when describing these two organs,
the physiological pH of the abdominal cavity of ver-
tebrates is maintained locally by very diverse home-
ostatic processes, and probably can persist at diverse
levels for a certain time after death. In addition, experi-
ments demonstrate that local shifts in the geochemical
gradients develop around a carcass as a result of mi-
crobial decay, too, and that these microbially-induced
shifts have a profound impact on whether or not miner-
als form and soft tissues are preserved (Briggs, 2003;
Noto, 2009; Sagemann et a/., 1999). Other experiments
have established that specific bacteria can play a key
role in phosphate precipitation. Some bacteria, includ-
ing the coliform bacterium Escherichia coli, appear to
facilitate phosphatisation by contributing phosphorus-
liberating phosphatases (Chin et a/., 2003; Hirschler e?
alM:990):

As soft tissues must be replicated before they are
destroyed by decay, rapid mineralisation was crucial in
Scipionyx. Of course, mineralisation relies on the avail-
ability of mineral ions. In the case of apatite, laboratory
experiments demonstrate that phosphorus may derive
from the decaying tissues of the carcasses themselves
(Briggs & Kear, 1993a; Noto, 2009) or may derive from
the surrounding sediment or pore water (Briggs, 2003).
These sources have been also deduced from the fossil
occurrences (Wilby & Briggs, 1997). For example, the
presence of fossilised soft tissue in several specimens
belonging to different taxa in the Solnhofen Limestone
reflects the presence of an external, supersaturated
source of phosphorus that is available to virtually all
carcasses. At Solnhofen, Briggs (2003) observed that
fossils with mineralised soft tissue tend to be covered
by a thinner lamina of sediment than those without, in-
dicating that they were closer to the sediment-water in-
terface, where adsorbed phosphorus was concentrated.

The Pietraroja Plattenkalk seems to postulate differ-
ent conditions. In addition to Scipionyx, other vertebrate
specimens are found to preserve soft tissue remains
(Costa, 1853-1864; D’Erasmo, 1914, 1915; Evans et ofip
2004), but many other fish and reptile specimens pre-
serve just skeletal parts. Some taxa never show any soft
tissue: decapod crustaceans, for instance, do not preserve
even traces of gills and muscle attachments (Bravi &
Garassino, 1998; Garassino, pers. comm., 2010), which
are instead commonly seen at Solnhofen (Garassino &

168 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Schweigert, 2006). At Pietraroja, phosphorus seems to
come from the carcasses, not from the sediment - at least
in the fossil Scipionyx. Our SEM element microanalyses
provide compelling evidence that all preserved body
parts, bone as well as muscle, cartilage and connective
tissue, are phosphorus-rich (Figs. 121A, C; 132B; 143C;
144A, C; 157D), whereas the embedding matrix is com-
pletely devoid of it. In fact, the samples of sediment
coming from the matrix around the specimen, including
the ones taken near the dinosaur’s outline and from the
same bed (Fig. 10), turned out to be entirely composed
of carbonates and silicates, with the former as primary
components (Figs. 13A, C). Therefore, as predicted by
Noto (pers. comm., 2009), the phosphorus of Scipionyx
is authigenic.

A survey conducted by Wilby & Briggs (1997) on
a range of Mesozoic and Tertiary laminated limestones
preserving fossils with phosphatised soft tissues, led to
the identification of three kinds of microfabrics. The
most spectacular preservation occurs where tissues are
phosphatised directly by very small apatite crystallites,
often <30nm: this produces a “substrate microfabric”
that may preserve details at the cellular or even subcel-
lular level. This is well-documented in Scipionyx by the
preservation of sarcomere-related banding in the caudo-
femoral myofibres (e.g., Figs. 157A-C), and indicates
that the carcass of the hatchling theropod was subjected
to very little decay and a rapid mineralisation in the pres-
ence of a high concentration of phosphates. This type
of soft tissue preservation is microbially induced, even
though the microbes themselves do not become min-
eralised (Briggs, 2003). According to Wilby & Briggs
(1997), substrate microfabrics tend to characterise fossil
fish, “the thick and toughened dermis of which provides
an effective barrier to microbes for a longer period”.
This statement leads us to infer that the squamous in-

Fig. 169 - SEM image of soft tissue remains and intestinal infill of the
duodenum of Scipionyx samniticus. The vacuolar aspect may be related
to microbial microfabrics (sensu Wilby & Briggs, 1997).

Fig. 169 - Immagine al SEM dei resti dei tessuti molli e del riempi-
mento del tubo intestinale a livello del duodeno di Scipionyx sam-
niticus. L'aspetto vacuolare dipenderebbe dalle cosiddette “microbial
microfabrics” (sensu Wilby & Briggs, 1997).

tegument of a dinosaur would provide a comparable, if
not tougher, barrier.

On the other hand, where tissues are easily perme-
ated and heavily invaded by microbes during the decay
of a carcass, the bacteria themselves may become the
site of apatite precipitation, and phosphatised microbes
may pseudomorph the gross form of the tissue. This
kind of preservation was described by Wilby & Briggs
(1997) as a “microbial microfabric”. Among the exam-
ples given, a microbial microfabric in a fish from the
Eocene of Monte Bolca (Wilby & Briggs, 1997: fig.
la-b) looks very similar to the one we see in all SEM
samples from the intestine of Scipionyx (Figs. 143A-B;
144B; 149A; 169; 170A). The infill of the intestine of
Scipionyx is dominated by a phosphatised (Figs. 143C;
144A, C), amorphous matter, characterised by large
concretions and many spaces. Notably, a very similar
texture, described as “spongy”, was found in the in-
testinal remnants of Mirischia (Martill et al., 2000). In
Scipionyx, this spongy, vacuolar aspect is widespread
in the mass embedding the soft tissue remains relat-
ed to the intestine, and some of the tissue themselves,
which are literally full of phosphatised microspheres
with an almost constant diameter that never exceeds
2 um (Figs. 144B; 170). This texture and the order of
magnitude of the microspheres are consistent with the
tissue having been replaced by pseudomorphed bacte-
ria, many of which are preserved as hollow spheres. We
infer from Wilby & Briggs (1997) that in Scipionyx,
contrary to somatic muscle tissue, such an extensive
microbial invasion of the visceral muscles occurred in
the intestine thanks to its nature, namely its structur-
ally high permeability. According to Wilby & Briggs
(1997), the detail preserved in a microbial microfabric
is related to the size of the bacteria, which are normally
1-2 um in diameter. Such an order of magnitude did not
affected preservation of the middle-gross anatomy of
Scipionyx, allowing us to observe, after more than 110
million years, the ripples of the mucosa of a dinosaur’s
intestine (Figs. 140-141).

ID-Xedal sd B
Window 008 - D9IS= IST

Fig. 170 - A) SEM image of a microsample of the rectum of Scipio-
nyx samniticus. The analysed microspheres (B), which in the intestine
of the Pietraroja dinosaur are very abundant (see also Figs. 143A-B;
144B; 149A; 169; 170A), could represent phosphatised bacteria. See
Appendix 1 or cover flaps for abbreviations.

Fig. 170 - A) immagine al SEM di un microcampione del colon di
Scipionyx samniticus. L'analisi delle microsfere (B), che nell’intestino
del dinosauro di Pietraroja sono molto abbondanti (vedi anche le Fig.
143A-B; 144B; 149A; 169; 170A), indicherebbe che si tratta di bat-
teri fosfatizzati. Vedi Appendice 1 o risvolti di copertina per le abbre-
viazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY | 69

In between these two kinds of preservation is the
“intermediate microfabric”, found where microbes are
not mineralised and much of the subcellular details are
lost. In this kind of preservation, the areas of tissue that
survive may preserve cellular traits, but where larger mi-
crospheres precipitate, they preserve only the coarsest
details. As an example of an intermediate microfabric,
Wilby & Briggs (1997: figs. 5b, 6a-b) illustrate muscle
fibres replaced by relatively coarse or large agglomera-
tions of apatite crystals, similar in aspect to the muscle
fibres we found in Scipionyx at the base of the neck. In
that area, in fact, the individual myofibres can be resolved
but the sarcomeres are more difficult to identify, as their
banding is faint (Fig. 132A, C) or lost (Fig. 132D).

Summing up, although substrate-, intermediate- and
microbial-microfabrics may occur rarely in the same
fossil, from what we can see in Scipionyx it is possible
that their gross distribution within the specimen reflects
the rate at which microbes gained access to the subcu-
taneous tissues of the organism. This phenomenon was
observed in other fossils by Wilby & Briggs (1997),
who regarded it just like the tendency of specific mi-
crofabrics to be associated with particular taxa and
indicating a taxonomic control that is likely to reflect
the rate of permeability and resistance to decay of the
integument of the different types of animals.

Diagenetic formations possibly related to soft tissues

Three calcite clusters embedded in the same plane of
the fossil may be regarded as being related to the decay of
the dinosaur carcass because of their proximity to the soft
tissue remains, their shape and their consistency, which is
definitely different from the surrounding, undisturbed ma-
trix (Fig. 117). The clusters or nodules differ from any other
fossil or sedimentary structure seen on the main slab in be-
ing homogeneously composed of calcite and in having a
eryptocrystalline aspect (Fig. 171), a grey-whitish colour
and a high compactness (Dal Sasso & Rampinelli, pers.
obs., 1994-2009). The transition to the fine-grained, yel-
lowish and softer micrite is abrupt, so that these nodules
have well-delimited margins.

The largest calcite nodule was considered in previous
studies (e.g., Ruben et a/., 1999; Hillenius & Chinsamy,
2004), and it is re-examined herein in a section devoted
to respiratory physiology, because of its improbable bio-

Fig. 171 - SEM image (A) and elemental spectrum (B) of the cranial-
most nodule of calcite associated with Scipionyx samniticus (Figs. 117,
172). See Appendix 1 or cover flaps for abbreviations.

Fig. 171 - Immagine al SEM (A) e spettro degli elementi (B) del nodulo
di calcite più craniale associato a Scipionyx samniticus (Fig. 117, 172).
Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

logical nature (therefore, we suppose it will become an
object of future debate). This nodule, 22.5 mm long and
4.1 mm wide, lies obliquely in the caudal half of the ab-
domen of Scipionyx. Of all calcite nodules, it appears to
be the deepest in cross section, and the one with the most
irregular surface, being composed of a cluster of large,
rounded granules, sub-spherical to drop-shaped and with
an average diameter of 0.7 mm each. This original aspect
is preserved all along the craniodorsal portion of the nod-
ule, but is lost in its ventral third, which has been polished
by preparation tools (Dal Sasso & Rampinelli, pers. obs.,
1997; see also Diaphragmatic Muscles).

A second calcite cluster, less extended than the former
(around 10.6 mm in diameter), is found between the right
humerus and the right radius, and delimited caudally by
the reddish halo derived from the blood-rich organs (Figs.
91, 172). This nodule seems quite thin, lying on the ex-
posed bedding plane as an almost flat, roughly subtrian-
gular area. It has a bumpy surface similar to that of the
former nodule, but is composed of granules that are more
irregular in shape, with an average diameter of 0.8 mm.

The third nodule, slightly more extended than the sec-
ond one but less exposed because it is hidden by many
bones, is visible ventral to and below the cranial dorsal
vertebrae and ribs: it is delimited cranially by the scapu-
lae, ventrally by the humeri, and caudally by the reddish
macula (Figs. 91, 172). Given its proximity to the second
nodule, and that it is covered by large bones and by the
reddish macula, we cannot exclude that this nodule might
represent a dorsal continuation of the second one. Because
of its position, the third nodule has been polished in order
to expose the bones (Dal Sasso & Rampinelli, pers. obs.,
1997), so that today it displays a smooth surface.

There is a fourth calcite area, which is not properly
shaped like a nodule, but is rather a vein that opens gradu-
ally between the right ceratobranchial I and the 9® cervical
centrum, then runs caudoventrally with an oblique, slightly
meandering course, widening to a width of 5.5 mm (Fig.
117). Along this 137 mm-long path, the vein crosses the
first two dorsal centra, the neck muscle remains, the left
scapular acromion, the right coracoid, the sediment and
finally the right manus. The bones crossed by the calcite
vein display some horizontal dislocation and vertical de-
formation (Fig. 172). This indicates that the vein formed
early in diagenesis, prior to sediment compression and
volume reduction. During this process of heavy compac-
tion, the rigid behaviour of the calcite vein should have
mechanically interfered in the nearest parts of the carcass,
provoking deformation of the bone and soft-tissue edges
where they were crossed by the vein, as both bones and
soft tissues acted in a more plastic way. According to Pez-
zotta (pers. comm., 2009), the syndiagenetic nature of the
calcite vein is revealed also by its sinous, non-angular ar-
rangement. In fact, this indicates that, when infiltration
of the calcium carbonate-rich solution occurred, the sedi-
ment still had a partially plastic behaviour. Nevertheless,
the shape of this vein and its distance from the dinosaur’s
body prevent us from inferring any relationship with any
anatomical structure.

The nature of the three nodules described above is dif-
ferent. According to Pezzotta (pers. comm., 2009), the
presence of nodules within the fossil and adjacent to it
cannot be fortuitous. The exceptional fossilisation in situ
of a variety of soft tissues with a chemical diversity in-

170 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 172 - Oblique view under grazing light of the thorax of Scipionyx samniticus, showing the syndiagenetic calcite vein (red arrow;
note the dislocated bones), the cranialmost nodule (black arrow), and the polished nodule (white arrow) that might represent a continu-
ation of the latter.
Fig. 172 - Vista obliqua e in luce radente del torace di Scipionyx samniticus, che mostra la vena di calcite sindiagenetica (freccia rossa,
notare le ossa dislocate), il nodulo in posizione più craniale (freccia nera) e il nodulo levigato (freccia bianca) che potrebbe rappresen-
tare una continuazione dello stesso.

herited from authigenic organic components indicates that
during diagenetic processes a rather low chemical mobil-
ity was maintained. Thus, as verified through recent ob-
servations of experimental taphonomy (Noto, 2009), the
chemical diversity of the decomposing tissues would have
perturbed and conditioned (in a sense, even polluted) the
local fluid chemistry, which generally is as homogenous as
a sedimentary layer devoid of inclusions, in the end cata-
lysing the formation of aggregates. That this phenomenon
occurred not only at the level of the dinosaur’s soft tis-
sues, but also in some areas of the sediment that directly
contacted the carcass, is proven by the formation of calcite
clusters both inside and outside the fossil of Scipionyx.

Likely, the formation of microcrystalline calcite aggre-
gates began at an early diagenetic stage. A concentrated solu-
tion of calcite salts accumulated in and around some parts of
the carcass because of the above-mentioned local chemical
(and probably, textural) inhomogeneities. Parallel to bone
and soft-tissue phosphatisation, calcite crystallisation took
place in the pseudo-nodules. According to Pezzotta (pers.
comm., 2009), chemical and structural inhomogeneities
would have conditioned even the crystal texture. One may
argue, for example, that the more globular texture of the
nodule enclosed in the abdominal area depends on a locally
stronger geochemical-structural disturbance. The collapse
of the carcass remains and the progress of diagenesis then
further deformed and squeezed bone, soft tissue and calcite
clusters against each other, rendering them almost flat. Con-
sequently, the calcite precipitation was syndiagenetic.

The calcite of the nodules does not replicate any or-
ganic remains; however, on the terms we set out above,
it can be related to their presence and, in some cases, po-

sition. For instance, there is reason to suppose that the
formation of the nodule cranial to the liver remnants was
promoted by the decay of the relevant amount of cartilage
composing the sternal plates. Sedimentary nodules are
sometimes found to wrap around fossil tetrapod girdles. A
well-documented example in the field (Arduini, 1993) is
represented by procolophonian parareptiles from Ranohira
(SW Madagascar). In that locality, a layer of Upper Per-
mian laminated shales contains dozens of cfr. Barasaurus
skeletons, that are fully articulated but preserved partly in
the layer and partly in nodules. To be more precise, the
pectoral girdle and the pelvic girdle of the same specimen
are included in two distinct nodules lying side-by-side,
with the rest of the skeleton lying in a stratum medial to
the nodules’ depth. This is in situ evidence that the forma-
tion of nodules is topographically related to girdle carti-
lages, the decay of which possibly produced denaturation
of chondroitin-4-sulphate, keratan sulphate and hyaluronie
acid, setting free groups with a negative valency. The latter
would have catalysed agglutination of positive ions within
the interstitial water that concentrated around carcasses in
the form of nodules during early stages of decay.

The nodules present in the cranialmost portion and in
the abdominal portion of the trunk of Scipionyx are sug-
gestive of the presence of anatomical virtual spaces, such
as lungs, airsacs and yolksacs. However, in the absence of
any evidence of anatomical structures in those areas, this
idea remains a pure speculation. In fact, as seen for the
largest calcite nodule (see Diaphragmatic Muscles), SEM
imaging documents that at the ultrastructural level the mi-
nor calcite nodules do not replicate soft tissue or any other
anatomical structure (Fig. 171A).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 171

PART III - FUNCTIONAL MORPHOLOGY AND PALAEOBIOLOGY

SKELETAL RECONSTRUCTION AND IN VIVO RESTORATIONS OF SCIPIONYX SAMNITICUS

Although the delicate skeleton of Scipionyx was sub-
jected to diagenetic crushing, it has remained beautifully
preserved and quite well-articulated (see Skeletal Tapho-
nomy). The presence of unfused skeletal elements, con-
sequent to the early ontogenetic stage of the individual,
favoured movement and rotation of the bones, preserv-
ing their original shape and limiting the degree of defor-
mation, as not yet tightly fused elements tend to move
with respect to one another and to lose contact with the
surrounding bones rather than undergo deformation. We
re-articulated the disarticulated elements, correcting both
diagenetic crushing and the slight distortion suffered by
the skeleton, in order to generate the skeletal reconstruc-
tions presented herein (Fig. 173). In this section, we will
list these corrections and point out the problems encoun-
tered during the reconstructions. The unpreserved charac-
ters (osteological or otherwise) represented herein were
inferred following the methodology proposed by Bryant
& Russell (1992), i.e., based on the cladistic distribution
of known features in related taxa (primarily the Compso-
gnathidae), when necessary choosing among equivocal
or different phylogenetic inferences on the basis of form-
function correlation and ecological affinities among the
taxa.

Cranial reconstruction - A 3D version of the skull
was first reconstructed from thin cardboard on the basis
of the (2D) anatomical plates (Figs. 34, 174, 175). The
model, scaled 4:1, helped us to verify the interpretation of
a number of features.

The skull of Scipionyx is almost undisturbed in lateral
view. The most relevant corrections were made at the level
of the frontoparietal suture and quadrate-quadratojugal

contact. Concerning the former, the rostral margin of the
parietal and the caudal margin of the frontal, diverging
laterally in the fossil, were realigned, giving a uniform
curvature to the cranial vault. The outline we obtained re-
sembles that of the ornithomimosaur Garudimimus, which
is preserved three-dimensionally (Kobayashi & Barsbold,
2005). The realignment of the parietal and of the frontal
also favoured a slight anticlockwise rotation of the postor-
bital region of the skull, which helped to restore the origi-
nal contact between the quadrate and the quadratojugal, as
clearly indicated by the presence of the quadratojugal facet
on the lateral surface of the former (see Quadrate).

The diameter of the scleral rings was reconstructed on
the basis of the left one, which is more complete. The es-
timated minimum number of 16 plates was taken as good
for the reconstruction (see Scleral Plates).

As for most theropods, the nasals, prefrontals and fron-
tals of Scipionyx must have been partly exposed in both
lateral and dorsal views. In particular, as mentioned in the
description, the prefrontal of Scipionyx is rather large, and
it probably formed, together with the lacrimal, a lateral
triangular projection in dorsal view. In this view, it is dif-
ficult to understand whether a parietal process of the pos-
torbital contacting the parietal was present or not: as the
preserved parietal does not show an unequivocal contact
surface or process for the parietal, the parietal process of
the postorbital was not reconstructed here.

As mentioned above, the ventralmost tip of the ptery-
goid-ectopterygoid contact, formed by the ventral projec-
tions of the caudomedial arm of the ectopterygoid and the
ventral bar of the pterygoid, projected from the roof of
the mouth and served to maintain the alignment of the
mandible.

—rr—-===——
at late e

pri i 7

Ni

Fig. 173 - Reconstruction of the skeleton of Scipionyx samniticus in dorsal and lateral view. Missing bones in light gray. Scale bar = 5 cm.
Fig. 173 - Ricostruzione dello scheletro di Scipionyx samniticus in norma dorsale e laterale. In grigio chiaro le ossa mancanti. Scala

metrica = 5 cm.

172

CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 174 - Cranial reconstructions of Scipionyx samniticus (line drawings). A) skull in ventral (palatal) view; B) skull in caudal (occipi-
tal) view; C) articulated skull and mandible with closed jaws in right lateral view; D) left hemimandible in medial view; E) mandible
in dorsal view; F) mandible in ventral view. Scale bar = 10 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 1 74 - Ricostruzioni del cranio di Scipionyx samniticus (disegni al tratto). A) cranio in norma ventrale (palatale); B) cranio in norma caudale
(occipitale); C) cranio e mandibola articolati, con bocca chiusa, in norma laterale destra; D) emimandibola sinistra in norma mediale; E) mandi-
bola in norma dorsale; F) mandibola in norma ventrale. Scala metrica = 10 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Fig. 175 - Cranial reconstructions of Scipionyx samniticus (shaded drawings). A) articulated skull and mandible in dorsal view;
B) articulated skull and mandible with open jaws in right lateral view; C) articulated skull and mandible with closed jaws in left ros-
trodorsolateral view. Scale bar = 10 mm.

Fig. 175 - Ricostruzioni del cranio di Scipionyx samniticus (disegni ombreggiati). A) cranio e mandibola articolati, in norma dorsale;
B) cranio e mandibola articolati con bocca aperta, in norma laterale destra; C) cranio e mandibola articolati con bocca chiusa, in norma
rostrodorsolaterale sinistra. Scala metrica = 10 mm.

173

174 CRISTIANO DAL SASSO & SIMONE MAGANUCO

With closed jaws, the upper dentition was visible
in lateral view and passed labially to the lower one,
which accommodates medially to the medial walls of
the upper tooth-bearing bones, as occurs in toothed
theropods (e.g., Weishampel ef a/., 2004).

In ventral view, the outlines of the jugals and max-
illae match both the reconstructed width of the palate
based on overall size of palatal bones (primarily pala-
tines and ectopterygoids) and the outline and position
of the dorsal skull roof and of the elements of the lat-
eral sides of the skull. As in many theropods, the pala-
tines assumed a steep position in our 3D model, the
vomeral processes becoming visible in lateral view
through the antorbital fenestra and probably forming
a vaulted palate.

The mandibular symphysis in ventral view was re-
constructed U-shaped, as suggested by the rostral bend-
ing of the left dentary (see Dentary, Skeletal Taphon-
omy). The retroarticular process, which is shorter than
in Sinosauropteryx (Currie & Chen, 2001) and Compso-
gnathus (Peyer, 2006), was reconstructed bulkier, too,
to match the preserved outline of the other post-dentary
bones. Finally, the reconstructed skull was articulated to
the neck in order to have the upper tooth row on the
horizontal plane (i.e., slightly flexed), a position useful
to show its full dorsal aspect also in the dorsal view of
the whole animal.

Axial skeleton - The neck of Scipionyx was recon-
structed by articulating the vertebrae with the centra
in maximum contact. This resulted in a dorsally de-
flected base of the neck, an almost vertical mid-cer-
vical region and a slightly ventrally deflected cranial
region (Fig. 173). This posture represents the typical
sygmoidal curvature visible in the reconstructed neck
of many theropods (e.g., Paul, 1988), which is based
on the offset cranial and caudal surfaces of the cervi-
cal centra. It represents neither the so called “osteo-
logical neutral pose”, as some of the vertebrae are not
articulated in maximum contact with the zygapophy-
ses, nor the usual orientation in vivo of extant amni-
otes, which usually hold their neck more extended and
their head more flexed (Taylor ef a/., 2009, and refer-
ence therein). Also, the reader must bear in mind that
in this reconstruction we obviously did not try to re-
store the unpreserved intervertebral cartilages, which,
according to observations on extant animals, enable
greater flexibility in the neck than the bones alone
suggest (Taylor et al., 2009, and reference therein).
However, some osteological evidence does give an
idea of the mobility of the neck of Scipionyx. In crani-
al and mid-cervical neural arches, the ventral bending
of the rostral half of the convex articular surfaces of
the prezygapophyses probably permitted great mobil-
ity (in particular flexibility) of the neck, which could
have been well-deflected ventrally when the animal
drank and explored the ground.

Concerning the dorsal vertebrae, no adjustments
were required, except for articulating the disarticu-
lated centra and their respective neural arches.

As already mentioned, the preserved curvature of
the shafts of the ribs permitted reconstruction of the
shape of the trunk in both dorsal and lateral views.
The cranial dorsal ribs have an almost straight mid-

dle tract, indicating a relatively narrow and laterally
flattened thoracic region; the caudal dorsal ribs are
evenly curved all along their shafts and, thus, delimit
an abdominal region more evenly rounded in coronal
section than the thoracic one.

The well-preserved gastralia mark the ventral bor-
der of the abdominal cavity, which is exactly aligned
with the distal margin of the pubis. Thus, the rib cage
and the gastralia give a reliable idea of the volume of
the body cavity. According to Claessens (2004), this
cavity could have been partly enlarged or restricted
by the musculature associated with the articulated
gastralia, obtaining a function analogous to the dia-
phragm of the Mammalia (see Gastralia, Respiratory
Physiology). The feasibility of these movements in
Scipionyx is supported by the morphology of the ar-
ticular surfaces of the medial gastralia, with the slid-
ing surface of the mvf articulating with the mdf of the
successive, counterlateral element (Fig. 85). A carti-
laginous sternum was also reconstructed, and articu-
lated to the ribs via the cup-like distal ends of the 3"
and 4' dorsal ribs.

Based on the most complete theropod fossils, the
tail of Scipionyx would have been definitely longer
than the presacral region, independent of the ontoge-
netic stage of the animal (e.g., Kobayashi & Lii, 2003).
Moreover, compsognathids are among the longest-
tailed theropods: Sinosauropteryx has the longest tail,
with an estimated number of 64 caudal vertebrae (59
of which are preserved), which is almost twice as
long as the presacral region, skull included (Currie &
Chen, 2001). We reconstructed the tail of Scipionyx
with 51 vertebrae, a number close to that (i.e., 49)
found in the complete tail of Sinocalliopteryx (Ji et
al., 2007a). Given that the vertebral centra of Huaxia-
gnathus and Compsognathus increase in length along
the tail up to the distalmost portion, where they start
to decrease (Hwang et al., 2004; Peyer, 2006), and
that the length of the centra in Sinosauropteryx in-
crease only up to the 6' caudal, then steadily decrease
(Currie & Chen 2001), we decided to reconstruct the
tail of Scipionyx maintaining an average length all
along the series, with the exception of a decreasing
distalmost portion.

Pectoral girdle and forelimb - The coracoids of
Scipionyx were reconstructed as being closely ap-
pressed. Their position can be inferred not only from
the taphonomical condition of the specimen (they are
found appressed one to the other), but also by articulat-
ing the epicleideum of the furcula with the acromion of
the scapula. A comparable reconstruction was obtained
for Coelophysis by Rinehart et al. (2007: fig. 3), who
demonstrated that the whole furcula closely conforms
to the cranial margin of the coracoids. The finding of 7y-
rannosaurus furculae (Larson & Rigby, 2005), together
with observations made on other tyrannosaurids (Ma-
kovicky & Currie, 1998) and on articulated skeletons
of other theropods (Barsbold, 1983; Chure & Madsen,
1996), confirm that the furcula could have articulated
with the acromion processes of the scapulae only if the
coracoids almost touched along the midline.

The thin distal end of the scapula, which is not com-
pletely preserved, was reconstructed as slightly expand-

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 1675

ed and squared, like in other compsognathids (e.g., Ji ef
al., 2007a: fig. 1). As the manus of Scipionyx could not
fold against the forearm in avian fashion on account of
the absence of a semilunate carpal (Carpenter, 2002), the
angle between forearm and manus was maintained as it
is in the fossil. The 1% digit was reconstructed as diverg-
ing from the 2° and 3" ones, according to the asym-
metrical condyles and its position of preservation, the
left manus of the fossil showing a mirror-like position
of all the three bones composing digit I with respect to
digits II and III. The pollex looks like it would have di-
verged from the second digit during flexion, as is the
case in most other theropods, instead of converging (1.e.,
being in opposition). Divergence would have increased
the spread of the fingers, permitting Scipionyx to hold
larger objects (Gishlick, pers. comm., 2000; Senter, pers.
comm., 2007). The suggestion of Currie & Chen (2001)
that penultimate phalanges that are longer than the an-
tepenultimate, together with deep, moderately curved
unguals that taper to sharp points, might indicate that
the hand was adapted for grasping, can be applied also
to Scipionyx.

Pelvic girdle and hindlimb - After re-articulating
and realigning the vertebrae, the ilium, which is fos-
silised in a slightly anticlockwise-rotated (about 6°)
position, was realigned in turn. The angle between
the pubis and the ischium resulted slightly increased,
from 45° to 60°, after we re-articulated the former to
the ilium, according to the shape of their peduncles of
contact (Fig. 173).

The difference in size between Sr4 and Sr5, with
the latter larger than the former, likely indicates that
in Scipionyx, as well as in many theropods, the iliac
blades diverged at both cranial and caudal ends when
seen in dorsal/ventral views, increasing their distance
from the vertebral centra.

Only the proximal portions of the tibiae and fibu-
lae of Scipionyx are preserved, so the length of the rest
of the hindlimbs can only be estimated. Varricchio ef
al. (2002) reported a significant negative allometry
of the distal hindlimb segments during growth in the
advanced maniraptoran 7roodon, indicative of long-
er-limbed hatchlings and juveniles. Tibiae and fibulae
are longer than femora in juvenile tyrannosaurids, but
the opposite occurs in Sinornithomimus (Kobayashi
& Lii, 2003). Only few data are available for comp-
sognathids. In Sinosauropteryx, the femur/tibia ratio
is 87% in the younger and smaller individual NIGP
127586, and 89% in NIGP 127587 (Currie & Chen,
2001). In the French Compsognathus the femur/tib-
ia ratio is about 83% (Peyer, 2006); in the German
specimen, that ratio is reported to be 76% (Hwang ef
al., 2004), although this individual is ontogenetically
younger than the French one. The femur/tibia ratio is
about 90% in Juravenator, Huaxiagnathus and Sino-
calliopteryx. As there are no fossils of adult Scipio-
nyx, and as the femur/tibia ratio is fairly conservative
within the Compsognathidae, with a slight increase
during ontogeny in those species represented by more
than one individual, we tentatively adopted the lowest
value recorded for compsognathids (i.e., 76%, in the
range of 76-90%) when reconstructing the hindlimb of
Scipionyx, taking into account that our specimen is an

early hatchling. Based on a similar reasoning, as the
range of metatarsus / femur ratios is 67-76% within
the Compsognathidae (e.g., Hwang et al., 2004), we
adopted the highest value for the metatarsus of Sci-

pionyx.

Body length and body mass - According to our
reconstruction, the approximate total length of this
specimen of Scipionyx did not exceed 50 cm. Contra
Martill et a/. (2000), the Scipionyx fossil is still fairly
three-dimensional despite the compression it sus-
tained. After having re-articulated its elements and
corrected for the deformation, the pelvis of Scipionyx
results comparatively slightly narrower mediolater-
ally than that of Mirischia, the three-dimensionally
preserved pelvis of which has a width approximately
30 mm wide across the sacrum. The outline of Scipio-
nyx°s body (solid black in Fig. 173) has been drawn
on the basis of the usual distribution and attachment
of the main muscular masses in theropods (e.g., Paul,
1988), adjusting for the size and shape of the bones
of Scipionyx, and for the bulk of the yolksac in hatch-
lings of extant archosaurs (Dal Sasso & Maganuco,
pers. obs., 2011). The opening of the cloaca has been
placed in proximity to the ischial feet, further ventral
to the position in which the preserved faecal pellet is
fossilised. If the position of the faecal pellet was con-
sidered indicative of the location of the cloaca instead
of that of the rectocoprodaeal valve, it would have
implied a base of the tail that was not thick enough
dorsoventrally, i.e., not consistent with the position
of the M. ilio-ischiocaudalis and M. caudofemora-
lis longus in Scipionyx and in other theropods (see
Pelvic And Hindlimb Muscles). A restoration of the
whole musculature was not attempted because this is
made extremely difficult to carry out by the early on-
togenetic stage of the specimen and the consequent
lack of clearly visible muscular scars and ligament
attachments, which usually develop later during ar-
chosaurian ontogeny (e.g., Brochu, 1996). However,
comparison with usual body proportions of small
extinct and extant coelurosaurs, including birds, per-
mitted us to tentatively estimate a weight in life of no
more than 0.2 kg.

Integument - Over the last two decades, several
specimens originating from China have provided di-
rect fossil evidence that the body of basal coelurosaurs
was extensively covered with “protofeathers” (e.g.,
Currie & Chen, 2001; Xu ef a/., 2004; Xu et al., 2006;
Ji et al., 2007a, 2007b). Unfortunately, the integument
of Scipionyx was not preserved (see Description, Ilio-
ischiocaudal Muscle Septa). However, based on the
results of our phylogenetic analysis, we think that Sci-
pionyx would have been protofeathered as well. The
pattern of distribution of the protofeathers on the body
of compsognathids is not yet clear: for example, the
mid part of the tail of Juravenator shows a scaly integ-
ument composed of small conical and non-imbricated
tubercles (Gòhlich & Chiappe, 2006). According to
these data, all but one of the palaeoartistic restorations
presented in the end pages of this monograph incor-
porate, to various degrees, a proto-feathered coverage
for Scipionyx.

176 CRISTIANO DAL SASSO & SIMONE MAGANUCO

GUT CONTENTS AND FEEDING CHRONOLOGY

As mentioned beforehand, a variety of allogenous
organic remains is found along the path of the digestive
tube of Scipionyx. Given their intimate relationship with
the fossil of the little theropod, these remains can un-
doubtedly be referred to ingested food items (Fig. 176,
see also Fig. 117). Among the osseous remains, the larg-
est are preserved in the stomach region; smaller bones
are contained in various portions of the intestinal tract.
In this section, we deal also with integumentary remains
that are found in the duodenum and in the rectum. Our
description follows a craniocaudal direction, which in
chronological terms (intake sequence) means from the
last meal to the first.

Oesophageal contents

At the level of the right scapular acromion, dorsal
to the tracheal tube and somewhat parallel to it, runs a
small and elongated cluster of tiny yellowish, flat, some-
times angular, shiny elements (Figs. 135, 177). Their
highly reflective surfaces are consistent with smooth in-
tegumentary structures, such as tiny reptilian squamae
or fish scales. In addition, other very small single ele-
ments, which based on their brown colour can be inter-
preted as bones, are present. At least one of them has
an arched shape and may represent a rib or a haemal
arch. Given that oesophageal peristalsis is relatively fast
(a matter of some seconds), it is statistically improbable
that Scipionyx would have swallowed food just before
its death. The most likely explanations for food being
found in the oesophagus would be (agonal) regurgita-
tion during death or the consequence of the relaxation
of the gastric sphincters after death — these dynamics
are consistent with clinical observations in modern birds

(Huchzermeyer, pers. comm., 2010). The mixture of the
oesophageal contents and the high degree of fragmenta-
tion support such an interpretation.

Worthy of note is that some scales/squamae along the
residual line ofthe oesophagus display a blackish pigmen-
tation, just like the squamae preserved in the intestine. A
similar pigmentation occurs also on the tips of the horny
claws of Scipionyx and, thus, leads to suspect some sort of
correlation in the mode or state of preservation: in other
words, it is possible that similar colours reflect analogous
chemical compositions (Schweitzer, pers. comm., 1998).
To the best of our knowledge, all integumentary deriva-
tives of extant terrestrial vertebrates are made of keratin.
As stated above, we do not possess any chemical analysis
from any portion of the Scipionyx fossil, but this subject
would merit deeper examination. In the holotype of Sino-
sauropteryx, for instance, the integumentary tufts cover-
ing the back and the tail are also dark, and very recently
they were found to contain fossilised melanosomes, just
like the feathers of birds (Zhang et al., 2010).

Gastric contents

A cluster of tiny bones, mostly hidden by the cranial
portion of the duodenum, is visible in the middle of the
thorax of Scipionyx (Figs. 176, 178A-B). Because they
look like the ribs of Scipionyx in gross position, colour
and diameter, at first glance most of these elements might
be reasonably referred to the dinosaur’s skeleton. How-
ever, careful examination finds these tiny bones to be su-
pernumerary and out of specific position with respect to
the dinosaur’s ribs; under the light microscope, more im-
portantly, they appear rather more complexly shaped than
the rod-like bone shafts of the ribs. Most likely, this clus-

a Lepidosaurian bones

MI Fish scales

EMI Lizard-like polygonal squamation

cu

DA

IA Fish vertebrae MI Bone and integument fragments

Fig. 176 - Map showing the distribution of the gut contents of Scipionyx samniticus.
Fig. 176 - Mappa della distribuzione del contenuto dei visceri di Scipionyx samniticus.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 1797/

Fig. 177 - Oesophageal contents of Scipionyx samniticus: close-up of some fragmented scaly/squamous elements (arrows). Scale bar = 0.5 mm.
Fig. 177 - Contenuto esofageo di Scipionyx samniticus: particolare di alcuni elementi scagliosi/squamosi frantumati (frecce). Scala
metrica = 0,5 mm.

ter of “extra” bones belongs to swallowed prey. In fact,
as stated above, the position of the cluster is immediately
cranial to, and left of, the duodenum of Scipionyx, con-
sistent with the position likely occupied by the dinosaur’s
stomach within the thoracoabdominal cavity.

Description - Four rod-like bones display side-by-
side aligned articular surfaces, with subrectangular blunt-
ended shapes. These are appressed to each other, just like
metapodial elements of most terrestrial reptiles. We re-
gard these articular surfaces as proximal, given that they
face tiny rounded bones reminiscent of mesopodials;
moreover, these probable metapodials are more consist-
ent in form, position and proportion with tarsals than with
carpals. Therefore, we interpret mesopodium and metapo-
dium as belonging to a still-articulated ankle and pes, with
a mediolateral width less than 3 mm.

In particular, we interpret the first three metapodial el-
ements, numbering them from the right 7" dorsal rib of
Scipionyx, as being metatarsals I, II and III. All of their
straight diaphyses disappear under the duodenum, to-
wards the centre of Scipionyx’s abdomen. Metatarsal IV
diverges slightly from the first three metatarsals, heading
towards the right humerus of Scipionyx; also, metatarsal
IV is the only one that, after passing under a larger bone,
reappears to expose its distal end. This seems to possess
a ginglymoid distal articulation, exposed in ventral view,
which would indicate that the whole ankle and pes are
exposed in this aspect and that they are right-side ele-
ments. A small bone portion emerges in front of the distal
articulation of metatarsal IV, and might belong to a still-
articulated phalanx 1 of digit IV. Proximal to metatarsal
IV, and seemingly in articulation with the possible dis-
tal tarsal 4, is a reniform bone, even more divergent than
metatarsal IV, which tapers distally to an oblique trunca-
tion. Its surface texture is similar to that of metatarsals
I-IV. This element might be an incomplete metatarsal V.
This is also supported by its peculiar shape and divergent
position, which are consistent with the ankle arrangement
of many lepidosaurian reptiles (e.g., Carroll, 1977: fig.
12: Cocude-Michel, 1963: figs. 4B, 21, 39C, 40C).

As the mesopodium is partially covered by the ribs
of Scipionyx, the exact number of its elements cannot be
counted. However, we tentatively identify the triangular
portion exposed in between the centrum of the 8" dor-
sal vertebra and the 7° right dorsal rib of Scipionyx as
a visible portion of an astragalus or tibia; the tiny bone
lying across the proximal end of metatarsals II and III as
a distal tarsal 3; the reniform portion of bone appressed
to the proximal end of metatarsal IV as a distal tarsal 4
or as the proximal articulation of a metatarsal IV; and the
bone area including the relatively large, flat, semicircular
element overlapped by the 7! right dorsal rib of Scipionyx
and surrounded by metatarsal V and distal tarsals 3 and 4
as a calcaneum plus at least one more tarsal.

The bone overlying the diaphysis of metatarsal IV is
characterised by the following morphology: a voluminous
cylindrical body, grooved by a shallow triangular depres-
sion; a hemispherical condylar end; a flattened portion,
arising from the central body towards the metatarsals
and inclined towards the condylar end; and a markedly
concave margin, opposite to the flattened portion, facing
the 6° right dorsal rib of Scipionyx. We regard this ele-
ment as a centrum and part of the neural arch of a procoe-
lous vertebra in left lateral view. In particular, the lack of
transverse processes, as well as its general morphology,
is compatible with a caudal vertebra just distal to the last
ones bearing transverse processes. The centrum has no
trace of an autotomic septum.

Another allogenous element is present between the
shafts of the 6° right and 7° left dorsal ribs of Scipionyx,
paralleling them and continuing under the right humerus
of Scipionyx. This long bone has a side with a shallow fos-
sa, elongated towards the midshaft, and a feebly concave,
asymmetric articular end facing obliquely towards the
same side. Thus, this bone might represent the proximal
half of an ulna, with an exposed articular surface for the
radius and a well-developed olecranon or, alternatively,
the distal portion of a fibula, with an exposed articular
surface for the tarsus.

Lastly, ventral to the right elbow of Scipionyx, in the
corner between the cranial loop of the duodenum and

I/:S CRISTIANO DAL SASSO & SIMONE MAGANUCO

Scipionyx's bones
Scipionyx's reddish halo / matrix
Scipionyx's intestine si,
Lizard-like polygonal squamation

Tibia or astragalus

Calcaneum or distal tarsal : 2
Distal tarsal 3 è)
Metatarsal V
Metatarsal |V, and ?distal tarsal 4 (*) di
Metatarsal Il 4)
Metatarsal Il A }
Metatarsal | E
Phalanx IV-1 9
|. Vertebra (?caudal) ;

MD ?Uina i:
BM Indeterminate bone

calcaneum

astragalus

Fig. 178 - Stomach contents of Scipionyx samniticus. A) line drawing of all exposed allogenous bones; B) close-up of the main cluster
of bones; C) comparison with the ankle of /guana (after Carroll, 1977). For further comparison, see also Derasmosaurus (Fig. 15).

Scale bar = 1 mm.

Fig. 178 - Contenuto stomacale di Scipionyx samniticus. A) disegno al tratto di tutte le ossa allogene esposte; B) particolare
dell’agglomerato di ossa principale; C) confronto con la caviglia di /guana (da Carroll, 1977). Per confronti ulteriori, vedi anche

Derasmosaurus (Fig. 15). Scala metrica = 1 mm.

the right ulna of the dinosaur, is a triangular bone patch
that, despite its small size, does not have anything to
do with the thin, delicate gastralia nearby. Given the
completeness of Scipionyx’s forelimbs, this bone can-
not possibly belong to the dinosaur. Although it is rela-

tively far from the main cluster of allogenous bones,
this element is found in the middle of the dinosaur’s
abdomen, at the same level as the gastric contents and
in a position where the long bones of the swallowed pes
are pointing.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 179

Taxonomic affinities - Several genera of reptiles
coeval with Scipionyx are known at Pietraroja. Unfor-
tunately, these taxa are represented by specimens that
are badly or poorly preserved in the skeletal elements
homologous to the bones swallowed by the dinosaur.
Therefore, we extended our comparison to include re-
lated forms from other localities.

Regarding the single vertebra, the combination of a
procoelous condition and the lack of an autotomic septum
in a caudal element, which is also devoid of a transverse
process, prevents us from ascribing it to Rhynchocephalia
(e.g., Derasmosaurus). Such a vertebra might, on the oth-
er hand, be referred to Mesoeucrocodylia (Pol & Norell,
2004), a clade of crocodilians which includes some taxa
with procoelous vertebral centra, such as Araripesuchus
and Crocodylia. Actually, in comparison with the vertebra
swallowed by Scipionyx, an indeterminate mesoeucroco-
dylian from the Late Cretaceous of Madagascar (MSNM
V5632) presents with a condyle which is less pronounced
and, more importantly, with a diameter definitely less than
that of its own centrum; in addition, its ventral margin
is not concave but straight due to the presence of two
carinae (Dal Sasso & Maganuco, pers. obs., 2008). Fur-
thermore, if the prey of Scipionyx was a crocodilian, the
very small size of its vertebra would force our attribution
to a rather tiny individual, i.e. a hatchling; unfortunately,
the vertebra does not show any neurocentral suture. Most
likely, this vertebra belongs to a small squamate lepido-
saur which lacked autotomic ability in the tail. This con-
dition is variably distributed, even at the generic level.
Among extant taxa, procoelous vertebrae are common
in all Squamata except Gekkonidae (Estes et a/., 1988;
Gauthier ef a/., 1988). Among basal forms and all forms
that cannot be ascribed to living taxa, lack of procoelous
vertebrae is reported in Bavarisaurus, Huehuecuetzpalli
and Scandensia, and in at least two genera sympatric with
Scipionyx: Costasaurus and Eichstaettisaurus (Evans et
al., 2004). Among the squamates from Pietraroja, only
Chometokadmon would be a possible candidate, but this
character cannot be checked in the type specimen MNP
539 (Dal Sasso & Maganuco, pers. obs., 2008). Among
exotic taxa, vertebral procoely is found in Ardeosaurus
and Hovalacerta as well as in paramacellodids (Evans et
al., 2004).

As for the hindlimb bones, the ankle with a non-
vestigial metatarsal V would exclude the most advanced
Crocodyliformes (e.g., Romer, 1966; Carroll, 1988), in-
cluding the Pietraroja taxon (Dal Sasso & Maganuco,
pers. obs., 2008 on SBA-SA 207240); it is, however,
consistent with the Lepidosauria (e.g., Cocude-Michel,
1963). Unfortunately, given the limited exposition of the
tiny foot, the potential for further comparison is limited
and does not allow taxonomic attribution to a lower hi-
erarchic level. For example, we can tell that metatarsals I
and IV seem slightly more robust than metatarsals II and
III, so that with respect to this character the ankle swal-
lowed by Scipionyx resembles a specimen (MPN A01/82)
tentatively ascribed to Rhynchocephalia by Evans et al.
(2004). On the other hand, we cannot ascertain if like the
squamate Chometokadmon and unlike the rhynchocephal-
ian Derasmosaurus (Dal Sasso & Maganuco, pers. obs.,
2008 on MPN $39 e MPN 541) the prey of Scipionyx had
very elongate metatarsals and phalanges, similar to those
of extant lizards.

In conclusion, based on verified morphological af-
finities we cannot refer all the bones preserved in the
stomach of Scipionyx to the same taxon with certainty.
However, we suggest that they belong to a single indi-
vidual (and, consequently, to a single taxon) because
they are clustered together, like in a single meal, and
because the shape and size of the vertebra and the possi-
ble ulna/fibula are compatible with the size of the ankle
(Cocude-Michel, 1963; Dal Sasso & Maganuco, pers.
obs., 2010 on MSNM specimens of extant and extinct
diapsid skeletons).

Intestinal contents

Following the winding tangle of the intestinal tube,
continuous changes are observed in the aspect of its sur-
face, which goes from smooth to bumpy and from opaque
to shiny. Most textural inhomogeneities are likely due to
ingested solid food, but where food is deeply embedded,
it is impossible to determine its nature. However, re-
mains surface in at least five points of the intestine (Figs.
176, 179, 180) and, thus, some preparation was possible
without damaging the intestine itself. Below we describe
these contents.

Two adjacent clusters of solid organic remains crop
out along the cranialmost side of the descending loop
of the duodenum (Figs. 178A, 179A-B). Both are well-
preserved and composed of several shiny, polygonal-
to-roundish elements, still partly connected and seem-
ingly arranged in non-imbricate contact. This pattern is
consistent with the integumentary squamation seen in
certain body regions of both extant and fossil Lepido-
sauria (e.g., Caldwell & Dal Sasso, 2004), so we are
confident in referring these remains to a closely related,
if not the same, taxon. In several elements of the dorsal-
most cluster, the individual edges are blackish in colour
(Fig. 179B). As explained above, this is reminiscent of
fossilised integumentary structures possibly having a
keratinous composition.

A third, apparently allogenous, element surfacing
along the descending loop of the duodenum is embedded
in a thin layer of longitudinal musculature, about 7 mm
caudally to the right elbow of Scipionyx (Fig. 179D). It is
a single tiny bone, coloured dark brown like the bones of
Scipionyx. Given its symmetric, asterisk-like shape, we
regard it as a fish vertebra in dorsal view. In fact, this is
the common aspect of single vertebrae of several bony
fish, which are frequently found scattered in many layers
of the Pietraroja Plattenkalk.

A fourth area of ingested remains is positioned in the
jejunum, adjacent to the cranial face of the 11'" dorsal cen-
trum (Fig. 179C). Here, a rounded, 3 mm-diameter cluster
of dozens of small cylindrical elements, seemingly devoid
of pronounced apophyses, can be seen partial sealed in the
intestinal wall: these are likely vertebral centra of a very
small fish (see Palaeobiological Significance Of The Gut
Contents of Scipionyx).

Two even smaller roundish fragments surface along
the straight tract of the intestine (jejunum-?ileum), ven-
tral to the centrum of the 1" sacral vertebra (Fig. 179E).
These are brown coloured, like all other fossil bones; they
are possibly very incomplete vertebrae, but their origin is
impossible to determine.

CRISTIANO DAL SASSO & SIMONE MAGANUCO

Fig. 1795 Intestinal contents of Scipionyx samniticus. A) polygonal squamation in the cranialmost duodenum; B) possible remnants
of keratin (arrows) on a second adjacent cluster of integumentary elements; C) cluster of vertebrae in the cranial jejunum; D) isolated
vertebra in the descending loop of the duodenum:; E).tiny bones (arrows) in the jejunal-ileal region. Scale bars = 0.5 mm.

Fig. 179 - Contenuto intestinale di Scipionyx samniticus. A) squame poligonali nel tratto più craniale del duodeno; B) possibili residui
di cheratina (frecce) in un secondo agglomerato, adiacente, di elementi tegumentari; C) ammasso di vertebre nel tratto craniale del
pg D) vertebra isolata contenuta nell’ansa discendente del duodeno; E) ossa minute nella regione digiuno-ileale. Scale metriche
= 0,5 mm.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 181

Fig. 180 - Rectal contents of Scipionyx samniticus: translucent scales embedded in a faecal mass. Scale bar = 0.5 mm. See Appendix
1 or cover flaps for abbreviations.
Fig. 180 - Contenuto rettale di Scipionyx samniticus: scaglie traslucide inglobate in una massa fecale. Scala metrica = 0,5 mm. Vedi
Appendice 1 o risvolti di copertina per le abbreviazioni.

Lastly, the rectum of Scipionyx contains a faecal pellet
with a number of tightly bound, scaly elements parallel-
ing the dorsal and ventral caudofemoral muscle bundles.
These elements have a blunt-cornered, sub-rectangular
shape; thin, flat edges and seemingly thicker centres; and
a highly reflective, yellow-orange semi-transparent sur-
face (Figs. 176, 180). They form 4 partially overlapping
rows composed of 4-6 units each; the total count of the
exposed scales can be estimated to be at least 17. The ar-
rangement in rows suggests that the dermal tissue con-
necting the scales had remained almost intact, but it is
also possible that the scales became secondarily bound
together within the faecal pellet — many more scales are
likely hidden within the rectum of Scipionyx. SEM im-
aging highlights that the exposed scales are composed in
their entirety of acellular lamellar bone (Fig. 181A-B),
characteristic and diagnostic of teleostean fishes (DeLa-
mater & Courtenay, 1974; Meunier, 1984; Bernardo de
Sant'Anna, pers. comm., 2010). According to Meunier
(pers. comm., 2009), the ornamentation of the scales is
reminiscent of the Osteoglossiformes. The shape of the
scales suggests that they are of the elasmoid type. At high
magnification, the surface ofthe scales appears acid-etched
by digestive enzymes. The presence of 9 growth lines on
the natural edge of the scale examined by SEM coincides
with the count resulting from the cross-sectioned broken
edges (Fig. 181C), documenting that the fish to which the
scales belonged had survived for 9 seasons (Bernardo de
Sant’ Anna, pers. comm., 2010).

Palaeobiological significance of the gut contents
of Scipionyx

Remains of ingested prey are not uncommon finds
in the thoracoabdominal cavity of theropod dinosaurs.
Among Compsognathidae, the first record came along
with the finding of Compsognathus longipes. This German
specimen was initially incorrectly described to harbour an
embryo (Marsh, 1881). It was then reported to contain a
prey (Nopesa, 1903), an interpretation that was confirmed
by Ostrom (1978), who identified the ingested animal as
a young individual of the basal squamate Bavarisaurus.
Bone remnants of several animals, possibly belonging
to sphenodontids or lizards, were found also within the
gastric area of the French Compsognathus (Peyer, 2006).
Recent finds in China confirmed the predatory abilities of
compsognathid theropods: a specimen of Sinosauropteryx,
for instance, was recovered with the remains of a lizard
skull in its thoracic cavity (Chen er a/., 1998); furthermore,
the holotype of Sinocalliopteryx contains parts of a dro-
maeosaurid leg (Ji ef a/., 2007a), and the abdominal cavity
of the holotype of Huaxiagnathus contains some rounded
objects that, given their texture and colour, are likely bone
chunks of a partially digested prey (Hwang et a/., 2004).

In the light of this well-documented record, Scipionyx
does not seem to add much to our knowledge. However,
at least four features render Scipionyx remarkably inform-
ative. First, preservation of soft tissues allows us to be
absolutely sure that the allogenous remains are situated

182 CRISTIANO DAL SASSO & SIMONE MAGANUCO

10HMm

Fig. 181 - A) SEM image of one of the fish scales found in the rectum of Scipionyx samniticus. Nine growth lines (arrows) can be
counted along natural (B) and broken (C) edges of the scale. Scale bar = 1 mm.

Fig. 181 - A) immagine al SEM di una delle scaglie di pesce trovate nel retto di Scipionyx samniticus. Sia lungo i margini naturali (B)
che lungo quelli fratturati (C) della scaglia, si possono contare 9 linee di accrescimento (frecce). Scala metrica = 1 mm.

within the digestive tube, so that they can be interpreted
unequivocally as ingested prey. Second, being trapped by
the gut walls, which became themselves fossilised, the in-
gested remains did not have the chance to move after the
dinosaur died. Consequently, the relative positions of the
ingested remains have enabled us to reconstruct a feeding
chronology for Scipionyx (Fig. 182). This is a remarkable
insight, impossible to determine in almost any other fossil
specimen. In fact, today we know that Scipionyx had fed
on at least five different prey. First, one or more teleoste-
an fish, that we estimate to be 4-5 cm long from the size
(0.9 mm) of the scales found. A second, certainly distinct
meal, was a smaller-sized vertebrate that, given the diam-
eters of the vertebrae (not exceeding 0.5 mm) clustered in
the jejunum, we presume to be a clupeomorph-like prey
of no more than 2-3 cm in length, after comparison with
the Clupavus sp. fossilised near Scipionyx (Fig. 8) and
taking into account a vertebral count of 40-50 elements
in similar teleosteans (Murray et al., 2005). A third prey
of Scipionyx was a lepidosaurian reptile covered with sq-
uamae that, based on their size (0.4-0.7 mm in diameter),
suggest an animal no more than 10-12 cm long (Dal Sasso
& Maganuco, pers. obs., 2010 on MSNM extant reptiles).
Fourth came the posterior leg and part of the vertebral col-
umn of another reptile, possibly a lizard. Based on Hoff-
stetter & Gasc (1969), a vertebral length of 2.4 mm and
an ankle diameter of 3 mm are indicative of a total body
length of 15-40 cm, a size half that of Scipionyx and that,
in all likelihood, would have prevented the Italian comp-
sognathid from swallowing this prey whole. In addition,
hunting a prey half its own size would have been a chal-
lenge too hard to face alone and, therefore, such a meal
is either indicative of parental feeding or of a scavenging
activity. Fifth and last prey items were smaller vertebrates
of uncertain affinities.

A third important feature that differentiates Scipionyx
from other theropod dinosaurs is that the Italian compso-
gnathid had a varied diet, similar to that of the spinosaurid

Baryonyx (Charig & Milner, 1986), consisting of fish and
terrestrial vertebrates. Irrespective of whether these prey
were captured by a hatchling acting as a primary, active
predator or by adult individuals providing food for ‘their
hatchlings, Scipionyx samniticus, as a species, clearly rep-
resents an ecological generalist. Just like the Solnhofen
Bavarisaurus, the lizard-like squamates of Pietraroja
were very fast running, agile prey. Therefore, in all likeli-
hood, Scipionyx, like Compsognathus, had keen eyesight
and was capable of rapid acceleration, high speed and
quick reaction and manoeuvrability. Fish might be caught
either alive along the shore, most likely while trapped in
tide pools, or dead and freshly stranded. Such behaviour
seems supported by the frequent finding of theropod-tram-
pled surfaces in shallow marine carbonate platform de-
posits: for example, in the Cenomanian tracksite of Sezze
(Latina, central Italy) facies analysis reveals continuous
changes from subtidal to supratidal conditions, and 180
small theropod footprints with multiple crossing orienta-
tions on a single surface testify that near-shore environ-
ments were often frequented by these animals (Nicosia
et al., 2007). Considering the relative richness in food of
such environments, it is likely that small-sized theropods
searched among beached algae for marine invertebrates or
stranded fish when the sea receded. Probably, like present-
day shorebirds they were adapted to come and go with
each tidal cycle. Such behaviour, as well as the variety of
prey revealed by the gut contents of Scipionyx, would im-
ply a high level of mobility, either for hunting or for scav-
enging. Given the early juvenile age of the specimen, this
fact strengthens the idea that it was fed, at least in part, by
adults (see also Remarks in Ontogenetic Assessment).

Last but not least, the well-preserved gut contents of
Scipionyx has offered insight into a fourth feature: the
number, variety and different localisation of ingested prey
has given us the chance to indirectly investigate some as-
pects of the digestive physiology of theropod dinosaurs.
This is discussed in the following chapter.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY | 8 p.

-—* Chometokadmon-like lepidosaurian
dea» Clupavus-like teleostean
ss \epidosaurian reptile

«Mo teleostean fish

t- ‘

__ 9.

@)
>»

Fig. 182 - The hypothesised prey of Scipionyx samniticus (drawn to scale), and sequence of intake (circled numbers), based on the gut

contents illustrated in Figs. 176-181. Scale bar = 5 cm.

Fig. 182 - Stima delle prede di Scipionyx samniticus (in proporzioni reciproche reali) e ordine di ingestione delle stesse (numeri cer-
chiati), basato sul contenuto dei visceri illustrato nelle Fig. 176-181. Scala metrica = 5 cm.

REMARKS ON THE PHYSIOLOGY OF SCIPIONYX

Digestive physiology

The intestine is the part of the alimentary tract that ful-
fils the dual task of digestion and absorption of nutritive
substances; both are among the most important metabolic
functions of an organism. As form and function are cor-
related with each other, the preservation of the intestine
of Scipionyx offers a remarkable opportunity for the study
of dinosaurian digestive physiology. However, studies on
extant vertebrates indicate that physiological processes
are influenced by a large number of parameters that can-
not be evaluated in a fossil, so the matter requires extreme
caution.

Certainly, inference from gross-anatomy is reliable
mostly when verified through the well-known, basic bau-
plans of extant taxa by comparative anatomy and veteri-
nary medicine. For example, the anatomy of the diges-
tive system of extant tetrapods has been demonstrated
to be primarily diet-related: the intestines of carnivorous
animals tend to be shorter and less complex than those of
herbivores, which must be longer and more specialised in
order to digest cellulose. This occurs in most taxa, includ-
ing reptiles (e.g., Skoczylas, 1978), birds (e.g., Ziswiler &
Farner, 1972) and mammals (e.g., Kardong, 1997), inde-
pendently of their metabolic behaviour. Consequently, the
short, wide intestine of Scipionyx indicates a high absorp-
tion rate, and it is an indirect sign of a diet based on animal
proteins, which are easily processed, but — unfortunately —
this cannot be used to infer endothermy or ectothermy. In
other words, an efficient metabolism is not the same thing

as a high metabolic rate. Efficiency, in terms of digestive
physiology, is an expression of how completely ingested
food is turned into energy for the production of new tis-
sue and the maintenance of all other metabolic functions
(McNab, 2002). Therefore, efficiency has nothing to do
with how quickly such conversions occur. In the case of
Scipionyx, we can expect a relatively short intestine to be
associated with a slow passage time of the food, in order
to allow absorption of most nutrients before the material
is egested. Of course the presence of dense, anastomosed
plications (plicae circulares) on the mucosa of the duo-
denum greatly increased the surface area of the intestine
of Scipionyx and gave an important contribution to its ab-
sorptional function.

Digestion and absorption times in Scipionyx were
probably also related to a number of extrinsic and intrin-
sic factors that, we repeat, cannot be evaluated in the fos-
sil. Skoczylas (1978) and Chin et a/. (2003), for example,
remark that in reptiles the rate at which food is retained
and processed is influenced by the frequency of the meals,
by the volume, kin, degree of comminution and chemi-
cal composition of the food, by the ambient and preferred
body temperatures, by the motility capabilities of the gas-
trointestinal tract and by the physical and mental state of
the animal. In extant tetrapods, especially in ectotherms,
temperature plays a relevant role, a lowered temperature
generally slowing down both time-of-passage and time-
of-absorption of nutrients. Diefenbach (1975), for in-
stance, recorded gastrointestinal passage times in Caiman
ranging from 99 hours, at 30°C, to 315 hours, at 1SSC%

184 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Under equal environmental conditions, the persistence
of food within the stomach depends on its nature. Hard
food resistant to gastric juice (acid-etching), such as bone,
may be retained for hours or days before being conveyed
to the duodenum. The good preservation of the lepidosau-
rian bones in the stomach region of Scipionyx (articulated
ankle; dissociated vertebra and long bone; no acid etch-
ing) suggests the meal was recent, probably 12-24 hours
before death (see Skoczylas, 1978).

In crocodiles, the pyloric opening of the stomach is
so small that it does not allow the passage of undigested
bones into the intestine: the faeces are entirely devoid of
bones, and, irrespective of the passage time of other food
items, bones remain in the stomach until fully dissolved;
other accidentally swallowed indigestible items (stones,
pieces of wood, nylon netting) remain in the stomach in-
definitely, unless they are regurgitated (Huchzermeyer,
pers. comm., 2011). An extreme case of stomach reten-
tion is that of “turtle” shells, which are found to be fre-
quently swallowed by American alligators (Janes & Gut-
zke, 2002). At ambient temperatures over the range of
30-35°C, the gastric retention time is demonstrated to in-
crease with the shell size, contrary to prior research claim-
ing that in reptiles larger meals cause shorter permanence
in the stomach. It has been suggested that in environments
where stones are not available for ingestion, turtle shells
may serve as substitute gastroliths, but a real advantage of
turtle consumption is the excellent source of calcium pro-
vided by the bony shell. This is confirmed by the observa-
tion that juvenile saltwater crocodiles, Crocodylus poro-
sus, prey extensively on crabs more than adults (Webb e?
al., 1991). In the calcium/phosphorus ratio, these authors
measured a decline from 7/1 in 30-60 cm long juveniles,
to 2/1 in 90-120 cm long juveniles, concluding that the
higher intake of calcium by the smaller animals is consist-
ent with their immediate need to enhance skeletal devel-
opment. Similarly, a Juvenile Scipionyx would have had
an intake of calcium higher than that of an adult, which
well explains the utility of fish: whole small fish are more
digestible than whole reptiles, birds or mammals of the
same size, because they contain thinner bones (Klasing,
1998). A second similarity that we may infer for Scipionyx
comes from data obtained by Webb et a/. (1991) on preda-
tor/prey size. The size of the prey eaten by wild, early
Juvenile C. porosus was found to be strongly bimodal
(i.e., a small number of large prey, mainly rats, and a large
number of small prey, mainly crustaceans). Similarly, the
prey eaten by Scipionyx consists of several small fish and
small reptiles, and only a single portion of a larger reptile
(Fig. 182).

Regarding solid, undemolished food, the digestive
tube of Scipionyx contains patches of squamose skin in
addition to bone remains. Many reptiles consume their
relatives, and some reptiles often eat their own sloughed-
off skin. Contrary to the calcium obtained from partial
bone digestion, keratin is a substrate that is difficult to
digest (Skoczylas, 1978), and nothing is known of what
might be the alimentary (nutritional) significance of kerat-
ophagy, even in extant reptiles. Osseous and chitinous ob-
Jects are hardly digested by reptiles. Consequently, with
the exception of crocodiles (see above), a large variety of
undigested material is released by reptiles through their
faeces. In snakes, this is mainly fur, hair, squamae, feath-
ers, bird beaks, claws, rattles of rattlesnakes, egg shells,

bone fragments, snail shells and insect remains (Scali,
pers. comm., 2010; Skoczylas, 1978); in varanid lizards,
the faeces may contain mammalian and reptilian teeth,
bones, egg shells, squamae and scutes, fish scales, crusta-
cean carapaces and claws, and insect cuticles (Blamires,
2004). Crocodiles possess particularly strong digestive
enzymes and dissolve bones entirely in the stomach
(Janes & Gutzke, 2002; Richardson et a/., 2000). In this
case, it is not the size of the meal but rather the size of the
swallowed particles, in particular bone, that determines
how long the food remains in the stomach: the larger the
pieces of bone, the more time is needed for dissolution
by the stomach acid (Huchzermeyer, pers. comm., 2011).
There is evidence also in mammals that particle size has a
large effect on digestive time and efficiency, but for rep-
tiles and birds, which cannot chew their food, this is a
constraint (Carrano, pers. comm., 2010): if bones, squa-
mae, scales, feathers and hair cannot be broken down in
the mouth, then they have not enough surface area to be
fully digested, even when retention time in the stomach is
prolonged.

The hatchling Scipionyx was not capable of chewing
bones. Therefore, the largest bones in the stomach must
have been broken up by the parents feeding it or were
broken already when picked up by Scipionyx. In the ex-
ceptionally “fossil-frozen” digestive tube of Scipionyx,
we see also that: (a) bones are not fully digested in the
stomach; (b) connective tissue connecting small bones
remains intact, at least in the stomach; (c) small bones
pass whole through the pyloric valve and remain identifi-
able in the duodenum and in the jejunum; (d) squamous
skin (keratine) survives at least to the duodenum; and (e)
osseous skin (enamel, dentin) reaches the rectum and, al-
though acid-etched, accumulates in the faecal pellet.

Given that caudal to the pylorus no further digestion of
bones takes place in the intestine ofextant vertebrates (e.g.,
Kardong, 1997), the bone fragments in the intestine of
Scipionyx must have left the stomach in the state in which
they arrived in the rectum (Huchzermeyer, pers. comm.,
2011). So, after partial gastric dissolution, osseous tissue
(bones and scales) swallowed by the hatchling theropod
was driven through the intestine and released with the
faeces. Similarly, keratinous tissue, reaching the intestine
in the form of intact patches, is expected to have been
mostly eliminated in the same manner.

Therefore, the digestive physiology of Scipionyx
seems more like that of extant lepidosaurs than that of
extant archosaurs. Crocodiles fully dissolve small- to me-
dium-sized bones and regurgitate large undigested bones
(Huchzermeyer, pers. comm., 2011); similarly, carnivo-
rous birds regurgitate bones and other hard prey items
within their pellets, producing faeces that are devoid of
solid inclusions, as well (Klasing, 1998; Crosta, pers.
comm., 2011). Possibly, theropod dinosaurs also regur-
gitated the largest chunks of bone, but the condition in
Scipionyx indicates that the swallowed bones were mostly
guided through the intestine, like in carnivorous mam-
mals and lepidosaurian reptiles, that have pyloric open-
ings larger than in crocodiles and produce faeces contain-
ing bones and other incompletely digested hard tissues.
This finding gives important confirmation to the suppos-
edly theropod origin of some bone-bearing coprolites de-
scribed in recent years (Chin et a/., 1998, 2003; Chin &
Bishop, 2007).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 185

It is well-demonstrated (e.g. Klasing, 1998) that extant
carnivorous and piscivorous vertebrates, by swallowing
whole prey, obtain a metabolisable energy reaching 75%
of the total available, leaving only 10% of that energy to
egested, and 15% to excreted undigested remains. Diges-
tion of bones, even if partial, is crucial to these animals
because the ratio of calcium to phosphorus in soft tissues
is very poor. The gut contents of Scipionyx indicates that
theropod dinosaurs, whether having less powerful diges-
tive enzymes than extant crocodiles or not, could take
advantage also from hardly digestible food items, even
without completely demolishing them.

Respiratory physiology

Except for a short tract of the trachea, there are no
remains or imprints of lungs, air sacs or other organs of
the respiratory apparatus preserved in Scipionyx. Despite
this, in recent years the size and position of the liver rem-
nants, as well as the purported presence of diaphragmatic
muscles, involved Scipionyx in a vivid, though sometimes
controversial, debate on the respiratory physiology of di-
nosaurs. Within the present detailed study, a clarifying
state of the art based on our direct observations is needed,
and probably expected by many readers.

Liver and diaphragm - Some authors found simi-
larities between the liver of crocodilians, which fills the
cranialmost portion of the abdominal cavity, and the soft
tissue remains of Sinosauropteryx (Ruben et al., 1997)
and Scipionyx (Ruben et al., 1999). In particular, based on
size and position of the possible liver of Scipionyx, which
under UV light appears particularly large, Ruben et al.
(1999) concluded that, like in crocodilians, in theropod di-
nosaurs the liver completely subdivided the visceral cav-
ity into distinct anterior, pleuropericardial and posterior
abdominal regions. Together with the purported remnants
of diaphragmatic muscles, this suggested that theropods
did not have avian-like lungs ventilated by air sacs, but
possessed bellows-like septate lungs that were ventilated,
at least in part, by a hepatic-piston diaphragm.

Unlike in mammals, the crocodilian diaphragm con-
sists of a sheet of non-muscular connective tissue that
adheres tightly to the dome-shaped cranial surface of
the liver, while lateral, dorsal and ventral aspects of the
caudal portion of the liver serve as sites of insertion for
the paired diaphragmatic muscles (Ruben ef al., 1997).
More precisely, the crocodilian diaphragm is composed of
a post-pulmonary septum, as well as a post-hepatic sep-
tum (Ruben et a/., 2003); the two, in association with the
hepatic-piston pump, augment the effectiveness of costal
aspiration (Perry, 1998).

According to Ruben et al. (1997, 1999, 2003), Jones
& Ruben (2001), and Chinsamy & Hillenius (2004), both
Sinosauropteryx and Scipionyx retain preserved outlines
of complete thoracic-abdominal separation, defined by
“a remarkably crocodilian-like vertically oriented parti-
tion coincident with the apparent dome-shaped anterior
surface of the liver”. We have not examined the specimen
of Sinosauropteryx personally, but the describers of the
holotype raised some doubts on that conclusion (Currie &
Chen, 2001). Careful observations made by Paul (2001,
2002) led him to think that in Sinosauropteryx, 60% of the

cranial edge of the carbonised material medial to the ribs
consists in breakage of the fossil slab, and the preserved
edge is irregularly formed. More remarkably, taking as
reference points homologous skeletal elements in Sino-
sauropteryx and Scipionyx, Paul (2001, 2002) noted that
the position occupied by the carbonised material in the
former is occupied by the intestine in the latter. So, above
the cranial end of the gastralia in Sinosauropteryx there is
no liver but an empty space, and the organic remains are
most compatible with traces of the stomach and/or intes-
tine.

Paul (2001, 2002) provided relevant information of
comparative anatomy also for Scipionyx. Among extant
tetrapods, livers tend to be largest in growing juveniles,
and larger in carnivores than in herbivores. Moreover, liv-
er size can vary within a given species: in some birds, the
liver is so large and tall that it almost spans the distance
from the sternum to the vertebrae, to the point that it even
extends up between the high-set lungs. Experiments on
the dynamics of yolksac resorption and post-hatching de-
velopment of the gastrointestinal tract in chickens, ducks
and geese (Jamroz et al., 2004) indicate that, during the
first 21 days of life, the liver and pancreas are the fastest
growing organs, increasing rapidly in size. So, the pres-
ence of fossil remains of a large liver that, even assum-
ing it as not compressed by post mortem burial, spans the
entire thoracic cavity and forms a cranially convex arch,
is compatible with either a crocodilian- or a bird-like ju-
venile internal anatomy.

Despite not having examined the fossil directly, Paul
(2001, 2002) correctly inferred that no septa are preserved
in Scipionyx. Our present study of the specimen did not
reveal any evidence of either post-hepatic or post-pulmo-
nary septa. Therefore, despite our agreement with Chin-
samy & Hillenius (2004) that no significant diagenetic
dislocation of the liver occurred, and even assuming that
the liver was as large as the halo produced by its decay,
one cannot demonstrate that it divided the body into two
cavities. We described the reddish matter as a halo (Dal
Sasso & Signore, 1998a; Dal Sasso, 2003, 2004; Dal
Sasso & Maganuco, this volume), because its shaded-off
margins are consistent with a liquid or decay-derived liq-
uefied material that contaminated surrounding elements
(including bones) after the animal died (Figs. 137-138).
By definition, such a halo would make it impossible to
discern any precise physical boundaries. In any case, al-
though the body of Scipionyx is more elongate than that
of a bird, the proportion of it taken up by the inferred
pleural cavity (Ruben ef al. 1998, 1999) is very small
and more bird- than crocodile-like (Huchzermeyer, pers.
comm., 2010).

One of the arguments used by the detractors of the
theropod-bird link relies on the observation that crocodil-
ians have a diaphragm and birds do not. Actually, some
authors (e.g., Ruben ef a/., 1997) argue that the earliest
stages in the derivation of the avian abdominal airsac sys-
tem from a diaphragm-ventilating ancestor would have
necessitated selection for a diaphragmatic hernia in taxa
transitional between theropods and birds. Such a “debili-
tating condition” (sic) would have immediately compro-
mised the entire pulmonary ventilatory apparatus and
seems unlikely to have been of any selective advantage.
We do not see any need of a hernia for a transition from a
double abdominal cavity to a single one. Rather, under an

186 CRISTIANO DAL SASSO & SIMONE MAGANUCO

evolutionary perspective we see the diaphragm more like-
ly gradually reducing in size and becoming segregated to
a small vestigial cavity, or changing its function, whilst
the avian lungs developed, becoming more and more ef-
ficient. Actually, Perry (2001) remarks that birds do not
lack abdominal septa: during ontogeny, in fact, the tho-
racic air sacs invade the avian embryonic postpulmonary
septum and divide it into two parts: the horizontal septum,
which forms the ventral aspect of the lung, and the oblique
septum, which abuts the liver. As mentioned above, the
diaphragm of crocodilians is partly formed by the post-
pulmonary septum. Thus, an evolutionary path similar to
the one that the avian ontogeny seems to recapitulate has
some probability to have been followed by theropod dino-
saurs. In any case, our opinion, corroborated by a recent
discovery (Farmer & Sanders, 2010; see below), is that
even in the presence of a complete diaphragm, primitive
avian-like lungs could have worked as well.

Diaphragmatic muscles - As mentioned above, lung
ventilation in crocodilians is achieved by costal aspira-
tion in association with a hepatic-piston pump, a derived
mechanism utilising a novel respiratory muscle, the M.
diaphragmaticus (e.g., Brainerd, 1999). This muscle orig-
inates from the caudal gastralia and cranial surface of the
pubes, and inserts on a collagenous fascia on the caudal
surface of the liver: contraction of this muscle pulls the
liver and viscera caudally, decreasing the pressure within
the thoracic cavity (Carrier & Farmer, 2000). In Scipionyx,
a calcified area surfacing just cranial to the pubic bones
was identified as the possible remnants of a diaphragmatic
musculature having an analogous mode of action to that
of the M. diaphragmaticus (Ruben et al., 1999; Chinsamy
& Hillenius, 2004). This assumption was based on com-
patibility of the position of the purported remnants with a
respiratory function, and on the observation that the sur-
face of that area showed, under suitable lighting condi-
tions (e.g., Fig. 185A), longitudinally oriented “imprints”
reminiscent of a craniocaudal arrangement of the muscle
fibres (Ruben et a/., 1999: fig. 3). Actually, our photo-
graphs reveal that this calcified area is part of a larger hard
nodule, which was partly polished during the preparation
of the specimen (Figs. 183-184; Dal Sasso, pers. obs.,
1997) and which continues into a bumpy cluster directed
craniodorsally in the abdomen, rather than horizontally
towards the supposed liver (Fig. 184). Moreover, the lon-
gitudinally oriented “imprints” of the nodular surface are
an artefact of preparation, produced involuntarily in 1997
by an MSNM collaborator (S. Rampinelli) when trying
to weaken and remove the very hard nodule. In addition,
comparative SEM imaging and microanalysis revealed
that the aspect and the phosphatic composition of the cal-
cite nodule definitely differ from those of the myofibres of
Scipionyx, even in the areas where the musculature is only
moderately preserved, such as in the neck region (Figs.
132, 185). Indeed, this nodule was found to be made in its
entirety of an amorphous, cryptocrystalline calcite, just
like the nodule found cranial to the right elbow of Scipio-
nyx (Fig. 171). Therefore, we conclude that the purported
remnants of diaphragmatic musculature are not consistent
with the state of preservation of all other muscle tissue
in Scipionyx and do not show any macroscopic evidence
(bundle-like appearance) or microscopic evidence (pres-
ence of myofibres) attributable to fossilised muscle. As a

Fig. 183 - This photograph of Scipionyx samniticus, taken in 1993, doc-
uments the original granular aspect of the nodule of calcite (red arrow)
contained in the abdominal region, just cranial to the pubic bones. See
Appendix 1 or cover flaps for abbreviations.

Fig. 183 - Questa fotografia di Scipionyx samniticus, datata 1993, te-
stimonia che il nodulo di calcite contenuto nella regione addominale,
subito cranialmente alle ossa pubiche, aveva in origine un aspetto gra-
nulare (freccia rossa). Vedi Appendice 1 o risvolti di copertina per le
abbreviazioni.

matter of fact, Huchzermeyer (pers. comm., 2010) points
out that the diaphragmatic muscle of crocodiles, albeit
strong, is structurally very thin. So, if it was present in
Scipionyx, it would probably not have been preserved or
would have been markedly and definitely thinner than the
purported diaphragmatic muscle remnants.

Lungs - Just as we were writing this section of the
monograph, a revolutionary discovery in comparative
anatomy was published (Farmer & Sanders, 2010). Up to
now, the lungs of birds were believed to be unique among
vertebrates. The avian respiratory apparatus is composed
of two main components — the rigid gas exchanging lungs
and the non-vascularised ventilatory air sacs — and air-
flow through most of the tubular gas-exchanging bron-
chi (parabronchi) of the avian lung is unidirectional, air
travelling in the same direction during both inspiration
and expiration; in contrast, in the lungs of mammals and,
presumedly, of all other vertebrates, air moves tidally into
and out of terminal gas-exchange structures, which are
cul-de-sacs. Unidirectional air flow in the avian appara-
tus was purported to depend exclusively on the bellows-
like ventilation of the air sacs, and was thought to have
evolved to meet the high aerobic demands needed for
sustaining flight. However, Farmer & Sanders (2010) dis-
covered that, despite the different external gross morphol-
ogy of the respiratory apparatus of birds and crocodilians,
the topography of the intrapulmonary bronchus and of the
first generation of bronchi is similar in these two taxa, and
that air flows unidirectionally through parabronchi in the
lungs of the American alligator. This strikingly bird-like
respiratory system functions without the presence of air
sacs and with a diaphragmatic breathing mechanism. This
finding eliminates all the problems raised so far with re-

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Fig. 184 - Oblique view of the nodule of calcite shown in Fig. 183, as it appears today. The ventralmost portion (red arrow) was pol-
ished during preparation in 1997, and then interpreted by some authors as remnants or imprints of diaphragmatic muscles. Actually,
its continuity with the craniodorsally directed portion, which retained the original granular texture (black arrow), is still evident. See
Appendix 1 or cover flaps for abbreviations.

Fig. 184 - Vista obliqua del nodulo di calcite mostrato in Fig. 183, come appare oggi. La porzione più ventrale (freccia rossa) fu levi-
gata durante la preparazione nel 1997 e in seguito fu interpretata da alcuni autori come residuo o impronta di muscoli diaframmatici. In
realtà la sua continuità con la porzione di nodulo diretta craniodorsalmente, che ha mantenuto la tessitura granulare originaria (freccia
nera), è ancora evidente. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

187

35 155 heV Oem ID=
A Vert=1000 Window 0.005 - 40955» 51918 cat

Fig. 185 - Photograph taken under very grazing light (A), SEM imaging (B) and SEM element microanalysis (C) of the purported
remnants of diaphragmatic muscles of Scipionyx samniticus. These analyses reveal an amorphous cryptocrystalline structure anda
composition of calcium carbonate, a result that is incompatible with the mode of preservation of muscle tissue in Scipionyx (F igs. 132,
157), but consistent with calcite nodules (Fig. 171). Scale bar = 1 mm. See Appendix 1 or cover flaps for abbreviations.

Fig. 185 - Scipionyx samniticus. Le fotografie in luce molto radente (A), le immagini al SEM (B) e la microanalisi degli elementi al
SEM (C) degli ipotetici residui di muscoli diaframmatici mostrano una struttura criptocristallina amorfa e una composizione in carbo-
nato di calcio, non compatibili con le modalità di conservazione degli altri tessuti muscolari di Scipionyx (Fig. 132, 157), ed equivalenti
a quelle degli altri noduli di calcite (Fig. 171). Scala metrica = 1 mm. Vedi Appendice 1 o risvolti di copertina per le abbreviazioni.

188 CRISTIANO DAL SASSO & SIMONE MAGANUCO

gard to the evolutionary transition from a crocodilian-like
to a bird-like breathing apparatus and, thus, represents a
fundamental advancement in evolutionary biology. It also
suggests that unidirectional pulmonary airflow dates back
to the basal archosaurs of the Triassic, in that A/ligator is
an ectothermic reptile without air sacs, and that this respi-
ratory modality may well have been present in all cruro-
tarsans as well as in ornithodirans, dinosaurs included.

Trachea - Ruben ef a/. (1999, 2003) and Chinsamy
& Hillenius (2004) drew attention to the ventral position
of the trachea in Scipionyx: the preserved tract of the tra-
chea, found in the caudal cervical region, is well-distant
from the vertebral column and immediately cranial to the
scapulocoracoid complex. These authors suggested that
this position was evidence of a crocodilian-like, dorsov-
entrally deep lung, contrasting with the location of the
caudal portion of the avian cervical trachea, which is usu-
ally more dorsal, i.e. adjacent to the ventral margin of the
vertebral centra, in order to facilitate entry of the trachea
into the dorsally attached parabronchi (McLelland, 1989).
Paul (2001, 2002) ascribed some ventral displacement
of the trachea of Scipionyx to the 10!" cervical vertebra
, but in our opinion this element is too distant, at least
from the preserved tract of trachea, to infer any cause-
and-effect dependence (Fig. 129). However, according to
Paul (2001, 2002), the position of the trachea in Scipionyx
is consistent with both crocodilian-like and bird-like res-
piratory systems. In fact, although it is common for the
trachea of birds to run just below the cervicodorsal verte-
brae, as described by Ruben er a/. (1999, 2003) and Chin-
samy & Hillenius (2004), in some birds, especially those
with long necks, the trachea runs more ventrally and is
remarkably mobile. Indeed, tracheal position may change
with the position of the neck at any given moment even in
the same individual, and the air passage may drop well-
below the vertebral column, just cranial to the shoulder
girdle (McLelland, 1989: fig. 2.8a). Also, this position is
not immediately adjacent to the lung entrance, so we tend
to agree with Paul (2001, 2002) that the fossil-frozen po-
sition of the trachea in Scipionyx does not offer definitive
evidence regarding lung depth.

Abdominal air sacs - In Scipionyx, the duodenum
seems to be slightly displaced cranially, while a more
anomalous displacement appears to have moved the rec-
tum caudally. These two parts of the digestive tract have
left a large, potentially empty space within the pelvic
cavity in front of the pubic bones. According to some au-
thors (Martill ef a/., 2000; Paul, 2001, 2002), this space
is suggestive of the presence of avian-like abdominal air
sacs. Contra Ruben et al. (1999: fig. 2; 2003: fig. 11),
the intimate attachment of the intestine to the vertebral
column in Scipionyx is not an evidence against the pres-
ence of air sacs, as in extant birds the intestine does not
have to hang loose in the abdomen to accommodate them
(Huchzermeyer, pers. comm., 2010). Although it is true
that the dorsal position of the rectum in Scipionyx is like
that of crocodilians and mammals, nobody has shown
to date that dorsally placed recta are not found in tetra-
pods that do not use the hepatic-piston pump (Paul, 2001,
2002). In any case, avian-like archosaurian lungs do not
necessarily require avian-like air sacs, as strikingly dem-
onstrated recently by Farmer & Sanders (2010).

Although carnivorous dinosaurs have pneumatised
bones, to date there is no direct evidence for theropod air
sacs in the fossil record. The only case in which some-
thing similar appears to be seen is that of the Brazil-
ian compsognathid Mirischia, the pelvis of which was
found well-preserved in three dimensions inside a nodu-
lar concretion from the Chapada do Araripe fossil site.
According to Martill ef a/. (2000), the large empty space
impressed in the sediment caudal to the pubic shafts was
left by an abdominal air sac. Such an air sac could have
been ventilated by a dorsal pneumatic duct passing be-
tween the sacrum and the pubes, and a ventral one pass-
ing through the distal opening in the pubic apron. How-
ever, observations on extant birds (Huchzermeyer, pers.
comm., 2010; McLelland, 1989) show that the walls of
air sacs are very thin, and that the air sacs of dead ani-
mals are only virtual spaces. Air sacs do not displace
abdominal organs cranially, and after death of the animal
they do not leave any evidence: as written by Paul (2001,
2002), “fossilising the air sacs of a bird is as improbable
as fossilising a balloon”. Also, the majority of the air
sacs lie lateral to the internal organs, so organ-free areas
will not necessarily mark the sites of air sacs in a later-
ally flattened specimen. In this respect, the alternative
yolksac hypothesis, that we discussed while describing
the rectum, fits better with the displacement of the in-
testine and the empty space seen in the abdominopelvie
cavity of Scipionyx.

Osteological correlates: pneumatic bones - Among
extant tetrapods, pneumatic postcranial bones are only
present in birds and are the osteological correlates of
the pulmonary air sacs. The air sacs serve as air stor-
age chambers and ventilatory bellows, which generate
airflow through the rigid lungs (Scheid & Piiper, 1989).
Extrapulmonary diverticula of the air sac system invade
the posteranial skeleton of birds, resulting in the pneu-
maticity of variable portions of both the axial and appen-
dicular skeleton (O°Connor, 2006). A recent study on the
avian vertebral column (Wedel, 2009) demonstrates that
this invasion takes place during ontogeny following a pre-
dominantly craniocaudal order, with diverticula of differ-
ent sources (cervical air sacs, lungs, abdominal air sacs)
pneumatising their respective skeletal domains at differ-
ent times. Wedel (2009) observed that this sequence in
bird ontogeny recapitulates the evolution of pneumaticity
in theropods and sauropodomorphs. In extant birds, Wedel
(2009) also found apneumatic vertebrae that are bordered
cranially and caudally by pneumatic vertebrae. These hi-
atuses are produced if the diverticula from the different
parts of the respiratory system do not meet. Similarly, the
presence of pneumatic hiatuses in dinosaurs suggests that
vertebral diverticula developed from different, specific air
sacs. Likely, as found in the postcranial skeleton of Ma-
jungasaurus (O°Connor & Claessens, 2005) and proven
in extant birds (Schachner et a/., 2009), the cervical and
cranial dorsal vertebrae of theropods were pneumatised
by diverticula of cervical air sacs; the middle dorsal verte-
brae were pneumatised by diverticula of the lung; and di-
verticula of the abdominal air sac pneumatised the caudal
dorsal, sacral and proximalmost caudal vertebrae.

Based on the distribution of pneumatic foramina in
the vertebrae (see Posteranial Axial Skeleton), Scipionyx
has at least two evident hiatuses: a cervicodorsal hiatus

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY | 89

and a dorsocaudal hiatus. The reason for the amplitude of
the pneumatic hiatuses in the axial skeleton of Scipionyx
could likely be the immaturity of the specimen, but what
is important here is that these hiatuses exist.

The diverticula pneumatising the skeleton of sauris-
chian dinosaurs might have evolved before they could
have become extra-osseous air sacs functional to lung
ventilation, producing useful exaptations like lightening
of the skeleton and the body mass. Under this perspec-
tive, postcranial pneumaticity may have facilitated the
evolution of unprecedented body sizes. In fact, the pneu-
matic diverticula of extant birds do not just replace bone
tissue, they also fill up space that in mammals are other-
wise occupied by fat. Sereno et al. (2008), for instance,
described several pneumatised gastralia in the tetanuran
theropod Aerosteon, suggesting that diverticula of the air
sac system were present also in surface tissues of the
thorax.

Osteological correlates: morphology of ribs and
vertebrae - The lungs of birds are rigid and do not change
volume during breathing. Each lung is positioned in the
craniodorsal region of the thoracic cavity, with the cos-
tal surface tightly attached to the ribs, and the vertebral
(medial) surface adhering to adjacent vertebral bodies
(Duncker, 1972). Because of this intimate connection, the
vertebral and costal surfaces of bird lungs are deeply in-
cised by the proximal section of each rib; as the ribs are
strongly bicapitate, composed of distinctly separate ca-
pitula and tubercula, approximately one-fifth to one-third
of the lung tissue is located in between the neighbour-
ing thoracic ribs (Maina, 2005). A recent report on living
and extinct archosaurs, studying thoracic rib morphology
and the position of the diapophysis and parapophysis of
dorsal vertebrae, suggests that most nonavian theropods
were similar to birds in that they possessed lungs that
were dorsally attached to the vertebral column and that
were deeply incised by the adjacent bicapitate thoracic
ribs (Schachner ef a/., 2009): this functionally constrained
the lungs as rigid, non-expansive organs that had to be
ventilated by nonvascularised accessory air sacs.

Like in most theropods, parapophyses and diapophy-
ses are completely separated along most of the dorsal
vertebrae in Scipionyx. Paralleling their vertebral articu-
lations, almost all the dorsal ribs are bicapitate, the ca-
pitulum and the tuberculum of each being well-distinct
(Figs. 60, 76A, 77). In crocodilians, on the other hand,
the parapophyses of almost all dorsal vertebrae are shift-
ed along the transverse process towards the diapophy-
ses, so much so that the two articulations form almost a
single surface. The result of this shift is a thoracic cavity
with a very flat and smooth ceiling caudal to the first two
ribs. According to Schachner e? al. (2009), this facili-
tates the cranial and caudal movement of the lungs when
inflated and deflated by the hepatic-piston pump. Based
on this, we remark that, as in most theropods, the man-
ner in which the bicapitate ribs of Scipionyx articulated
with their corresponding vertebrae generated a rigid
ribcage around lungs with a limited mobility compared
to those of crocodilians. Moreover, the corrugated ceil-
ing resulting from this arrangement of the axial skeleton
would have greatly inhibited the infiation of the lungs
by a crocodilian hepatic-piston mechanism. This mecha-
nism likely would not function with lungs that were in-

cised and fixed in place by the thoracic ribs and, thus, the
axial osteology of Scipionyx and most nonavian thero-
pods contradicts the hypothesis that they may have had a
hepatic-piston pump mechanism of ventilation.

Osteological correlates: gastralia - Because the com-
plex rib-sternum complex seen in birds did not evolve in
the nonavian theropods, Schachner et a/. (2009) hypoth-
esised that the protoavian lung must have been ventilated
in a different manner, possibly by gastralial or pelvic aspi-
ration. In contrast to the reduction of the gastralia seen in
other amniote groups, theropod gastralia show elaborate
modifications. The imbricating articulations observed in
theropods, including the peculiar morphology we de-
scribe for the gastralial heads of Scipionyx (Figs. 82, 85;
see Gastralia), unite the individual gastralia into a single
functional unit. As remarked by Claessens (2004), no sig-
nificant movement of individual gastralia can take place
without affecting the position of other components. The
mid-ventral articulations limit movementto a single plane.
Effectively, retraction and protraction of the gastralial
system narrows and widens the ventrolateral dimensions
of the theropod trunk. Thus, the anatomy of the gastral-
ial system in theropod dinosaurs indicates a more active
function than merely abdominal support or protection. A
plausible function for active retraction and protraction
of the gastralial system is lung ventilation. According to
Claessens (2004), the gastralia may have functioned as an
accessory component of the aspiration pump, increasing
tidal volume. Moreover, if the caudal region of the lungs
in some theropods had differentiated to form abdominal
air-sacs, the gastralia might have ventilated them. Gastral-
ial aspiration may have been linked to the generation of
small pressure differences between potential cranial and
caudal lung diverticula, which may have been important
for the evolution of the more specialised unidirectional
airflow lung of birds.

Skeletal adaptations consistent with an avian-like as-
piration pump are already present in basal neotheropods
(O’Connor & Claessens, 2005). In the thoracic ribcage,
the vertical arrangement of the diapophyses and para-
pophyses ensures a rigid and relatively incompressible
skeletal framework around the pleural cavity. As already
stated, also in Scipionyx the orientation of the vertebro-
costal articulations changes to a horizontal position in the
caudalmost elements. This shift allows lateral excursion
of the last dorsal ribs, which, along with movements of
the gastralial apparatus, may provide relevant volumetric
changes in the caudal half of the trunk.

According to Sereno et al. (2008), the evolution of a
primitive costosternal pump took place only in manirap-
toriform theropods with osteological correlates that Sci-
pionyx lacks, such as ossified sternal ribs and sternum,
and specialised joints between vertebral ribs, sternal ribs
and sternum. Hypapophyses and uncinate processes are
other key osteological characters used as evidence for a
more advanced protoavian respiratory system (Codd ef
al., 2007).

Osteological correlates: pelvic girdle - The hepat-
ic-piston hypothesis, suggested by some authors for the
theropod respiratory physiology and partly discussed
above, relies also on controversial osteological homolo-
gies in the pelvic girdle, which led to putative similari-

190 CRISTIANO DAL SASSO & SIMONE MAGANUCO

ties between theropods and extant crocodilians (Ruben
et al., 1997, 1998, 1999). Actually, the pelvis of croco-
dilians is highly derived and modified in relation to the
basal archosaurian condition in that the pubis articulates
only with the ischium, forming a mobile joint that is cor-
related with the action of M. diaphragmaticus (Carrier &
Farmer, 2000). On the other hand, the pubes were certain-
ly immobile in theropods (Scipionyx included) and most
other non-crocodyliform archosauromorphs. According
to Hutchinson (2001), this feature weakens the inference
that any theropods had such a ventilatory mechanism. Ru-
ben et al. (2003) replied that experimental data falsify this
assertion, because alligators with surgically fixed pubes
exhibit no alteration in intra-abdominal pressures during
exercise-induced periods of enhanced lung ventilation,
although caudal, inhalatory rotation of the liver and lung
tidal volume are reduced by about 15% in these animals.
Hutchinson (2001) casted doubt on the inference of
M. diaphragmaticus in any noncrocodiliform archosaur,
arguing that the crocodylian pubic foot is a vestige of the
ancestral archosauriform pubic apron, whereas in teta-
nuran theropods the pubic foot is an expansion of the lat-
eral surface of the distal pubic symphysis. Furthermore,
he remarks that Ruben et al. (1997) compared crocodil-
ian pubes with theropod pubes in lateral view, but these
are not homologous lateral bone surfaces because the
crocodilian bone surface shown is caudal, not lateral,
and is occupied primarily by the origin of the M. pu-
boischiofemoralis externus, not the M. diaphragmaticus
(Hutchinson, 2001: fig. 14). Ruben et al. (2003) replied
that in crocodilians the pubic rami are not major sites of
origin for the diaphragmatic musculature; instead, the
large, ventral portion of M. diaphragmaticus attaches
primarily to the last pair of gastralia; some portions of
this muscle may originate directly from the pubis, but
these attach to the craniolateral edge of the distal part of
the pubis, ventrally adjacent to the sites of attachment of
the puboischiofemoral muscles. Whereas the presence of
diaphragmatic muscles in theropods remains to be dem-
onstrated, it does not preclude air sac development, as
recently observed by Farmer & Sanders (2010) and first
discussed by Perry (2001). The latter author found that
hypothesis particularly interesting, because under the
perspective of a costal breathing mechanism aided by ac-
tively kinetic gastralia (Claessens, 2004), the main func-
tion of a diaphragmatic-like muscle, if present in a pre-

avian theropod abdomen, could have been to fix the liver
dynamically during inspiration and, thereby, to increase
the efficiency of costogastralial ventilation. According
to Perry (2001), a ventrally oriented contraction would
prevent dorsal movement of the liver, assuming it to lie
ventral to the lungs, as in birds and in fetal crocodilians,
and as we infer from Scipionyx. It would also explain the
presence of the post-hepatic septum as a relic structure in
birds. This septum stabilises the liver piston and attaches
the liver to the dorsal body wall in crocodilians, but lacks
a compelling function in birds.

Nevertheless, a significant flaw in a theropod dino-
saur with a hepatic-piston mechanism identical to that
of extant crocodilians, as hypothesised for Scipionyx and
Sinosauropteryx (e.g., Ruben et al., 1999; Chinsamy &
Hillenius, 2004), concerns the basic biomechanics of bi-
pedal locomotion. All theropods are obligatory bipeds,
whereas all crocodilians are obligatory quadrupeds. When
the abdominal viscera shift cranially and caudally by the
action of the diaphragmatic muscles in crocodilians, the
centre of mass also shifts (Farmer, 2006). This is not a
problem for a quadrupedal sprawling animal with a low
centre of mass, but it may create an equilibrium problem
for a parasagittally erect biped by shifting of the centre of
mass cranially and caudally with every breath, disrupt-
ing balance and agility (Schachner et al., 2009). Despite
the arguments for the presence of a hepatic-piston based
respiratory apparatus in theropods, both the biomechan-
ics of bipedalism and the osteology of the theropod axial
skeleton make this hypothesis extremely weak from a
functional point of view.

Finally, we like to repeat a conclusive remark made
by O’Connor & Claessens (2005): “recent studies of
non-avian theropod dinosaurs have documented several
features once thought solely to characterize living birds,
including the presence of feather-like integumentary spe-
cializations (Xu et al., 2004), rapid, avian-like growth
rates (Padian ef a/., 2001; Erickson et a/., 2001), and even
bird-like behaviours captured in the fossil record (Norell
et al., 1995; Xu & Norell, 2004). Either implicitly or ex-
plicitly, these studies have linked anatomical, physiologi-
cal or behavioural inferences with an increased metabolic
potential, suggesting that if not bird-like in metabolism,
theropods were at least “more similar” to birds than to
reptiles”.

CONCLUDING REMARKS

Scipionyx samniticus represents not only the most
complete predatory dinosaur from the middle to Late
Cretaceous of Europe, but also the dinosaur with the best
preserved internal anatomy ever found so far in the world.
It is a key fossil specimen, which deserved a long and
detailed examination. We did our best to accomplish this,
but we are aware that some subjects could be more thor-
oughly studied. For example, even though the whole os-
teology of the specimen has been revised bone by bone,
addressing our knowledge of the anatomy of the species
and of the Compsognathidae as well, it remains difficult
to determine to what extent the basal phylogenetic posi-
tion of Scipionyx within the Compsognathidae is due to
its generally plesiomorphic condition or to its immaturity.

Further finds, desirably including a growth series of the
species, are needed to support any postnatal trajectory in
Scipionyx and to improve our understanding of what can
be observed in other compsognathids as well as of the
phylogeny of the group.

New skeletal reconstructions of the holotype of
Scipionyx samniticus have been made based on its over-
all skeletal anatomy. Besides skeletal reconstructions,
a gallery of artwork by several Italian palaeoartists has
been included herein, livening up the end pages of this
monograph.

A summary of the biodiversity of the coeval flora and
fauna from the Lower Cretaceous (Albian) of Pietraroja,
as well as a short review of the present knowledge about

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 191

the depositional environment of the Pietraroja Plattenkalk
and the paleogeography of southern Italy at that time con-
firmed that Scipionyx samniticus inhabited a coastal envi-
ronment under a tropical-subtropical climate.

The preservation of a variety of ingested allogenous
remains in the gut of Scipionyx has provided a rare
glimpse of trophic biology in a Lower Cretaceous ecosys-
tem. Based on the nature of those animals, we now have
compelling evidence that Scipionyx fed on both lizards
and fish, with the latter being regular, non-occasional,
meals. This confirms and extends a hypothesis by Naish
et al. (2004), who found ecological similarities between
the Solnhofen and the Santana lagoons, with a variety of
theropods, including compsognathids, acting as general-
ist carnivores, scavenging at the shoreline and fishing in
shallow waters. Given how immature the holotypic indi-
vidual of Scipionyx was at the time of death (probably
less than three weeks old, maybe only three days old), it is
fair to state that the presence of gut contents might be evi-
dence of parental care and feeding. Parental care means
protection of the offspring (in birds it also includes keep-
ing the hatchlings warm) and guiding to food sources, 1.€.,
teaching the offspring how to obtain food and which food
to eat; parental feeding implicates also provision of bits
of food.

In any case, the most striking result ofthe present study
is that it documents a unique variety of fossilised soft tis-
sues, which at present renders this specimen of Scipionyx
the most remarkably preserved dinosaur known, the Li-
aoning feathered examples notwithstanding. Investigation
of the taphonomy of the specimen led to the conclusion
that the dinosaur’s carcass was exquisitely preserved be-
cause it had undergone very little decay and a rapid min-
eralisation process in the presence of high concentrations
of phosphates. Through detailed examination under opti-
cal microscopy, ultraviolet-induced fluorescence photog-
raphy and scanning electron microscopy coupled with el-

ement microanalysis, we have demonstrated that the soft
tissues of Scipionyx samniticus are mineralised in three
dimensions, and that their preservation is exceptional
even at cellular and subcellular levels. We presume that
future examinations of Scipionyx with more advanced,
hopefully non-invasive, techniques will uncover further
remarkable information.

Perhaps the most relevant results in terms of the con-
tribution to the debate on dinosaur physiology have arisen
from new studies on the reddish halo preserved in the tho-
rax of Scipionyx and on the “diaphragmatic muscles” pur-
ported by some authors. That the reddish halo is referred
to the liver, and likely also to the decay of the heart and
the spleen, has been confirmed by its haematic origin, but
unfortunately there have been no insights into either the
original structure of the organ or to a link to any sort of di-
aphragm. The remnants of “diaphragmatic musculature”,
on the other hand, turned out to be an amorphous calcite
nodule with a structure that is inconsistent with that of
the preserved muscles in Scipionyx. So, the hypothesis of
a hepatic-piston-assisted breathing mechanism in thero-
pods does not receive any support from this study of the
Italian compsognathid.

We think that the present description and the sup-
porting figures, allowing the comparison of the soft-tis-
sue morphology from a relevant extinct taxon with the
analogous biological structures of extant vertebrates,
may interest a broad community of scientists, includ-
ing palaeontologists, evolutionary biologists, functional
morphologists, comparative anatomists, biologists, vet-
erinaries, herpetologists and ornithologists. In addition,
given the popular appeal of dinosaurs, we also feel that a
non-scientific audience will be interested in such unusual
“palaeo-autopsy”, as well as the media will be fascinated
by the amount of information provided by such a tiny,
single fossil specimen on the life and death of a hatchling
predatory dinosaur.

ACKNOWLEDGEMENTS

We wish to thank all the people who helped us in the
study of Scipionyx, especially in the last 4-5 years. Dur-
ing this period, five superintendents succeeded to the head
of the present Soprintendenza per i Beni Archeologici di
Salerno, Avellino, Benevento e Caserta: Giuliana Tocco
(1986-2007), Angelo Maria Ardovino (2007-2008), Mario
Pagano (2008), Maria Luisa Nava (2008-2010) and Adele
Campanelli (2010-present). We are grateful to all of them
for having always met our needs in the examination of the
fossil Scipionyx, ensuring the continuity of our research.
Continuity was guaranteed also by the tenure of Leonardo
Vitola, the person responsible for the Department of Pho-
tography. In photographing Scipionyx over the years, he
got very keen on any new finds, just like the authors of
this monograph, and, thus, contributed greatly in increas-
ing the value of this unique fossil specimen.

Without the sharp eye of the person who collected the
tiny specimen in the Pietraroja quarry, nothing of what
is documented herein could have been imagined. Gio-
vanni Todesco, but also those who first recognised the
scientific meaning of the first Italian dinosaur (Giovanni
Pasini and Giorgio Teruzzi), paved the way with crucial
decisions and actions. The 1993 editorial staff of Oggi

magazine (Paolo Occhipinti, Pino Aprile and Franco Ca-
pone) helped the return of the specimen to the competent
authorities. Sergio Rampinelli performed a superb prepa-
ration of the fossil from 1994 to 1997. Giuseppe Leo-
nardi and Marco Signore were former fellow travellers
with one of us (CDS). In more recent years, Mariella del
Re and her collaborators at the Museo di Paleontologia,
Università di Napoli “Federico II”, provided access to
the historical specimens collected at Pietraroja by Costa,
D’Erasmo and others.

Three referees patiently read and reviewed this mono-
graph according to their field of research: Karin Peyer
(Muséum national d’Histoire naturelle, Paris), an expert
in fossil compsognathids, focused on osteology; Matt
Carrano (Smithsonian Institution, Washington DC) kindly
commented on soft tissue anatomy, physiology and tapho-
nomy; and Fritz Huchzermeyer (Onderstepoort Veterinary
Institute) offered his valuable help with a truly “fresh”
observational perspective with respect to traditional ver-
tebrate palaeontology — as a veterinary pathologist, Fritz
has dissected thousands of crocodiles and ostriches, ex-
periencing in the best way the comparative anatomy of
extant archosaurs.

192 CRISTIANO DAL SASSO & SIMONE MAGANUCO

At the end of the review process, Mike Latronico
quickly revised our (too approximate) English syntax. We
are also grateful to our colleagues Anna Alessandrello and
Michela Mura for the editing of this huge volume, and to
Claudio Pagliarin and Graziella Perini for scanning many
photographs.

Fundamental technical support, which greatly im-
proved our knowledge of Scipionyx, was given by the fol-
lowing persons and institutions:

Michele Zilioli (Laboratorio di Microscopia Elettroni-
ca, MSNM) operated the SEM for hours and captured all
the SEM images and SEM element microanalyses pub-
lished herein; Paolo Gentile (Dipartimento di Scienze
Geologiche e Geotecnologie, Università degli Studi di
Milano-Bicocca) gold-coated the microsamples for SEM;
Roberto Appiani (Gruppo Mineralogico Lombardo) took
all the photos under UV light and many other photos
under visible light; Armando Cioffi (Computerized To-
mography, Healthcare Sector Imaging, Siemens S.p.A.,
Milano) optimised the CT scan images, after operating
with the equipment given at our disposal by Piero Bion-
detti (U.O. Radiologia, Fondazione Ospedale Maggiore
IRCCS, Milano); Stefano Torricelli (SPES, Biostratigra-
phy & Sedimentology, ENI S.p.A., Exploration & Pro-
duction Division, San Donato Milanese) and Paola Maf-
fioli (Laboratorio di Micropaleontologia e Paleoecologia,
Dipartimento di Scienze Geologiche e Geotecnologie,
Università degli Studi di Milano-Bicocca) searched for
microfossils in our sedimentological samples; Fabrizio
Rigato and Maurizio Pavesi (Sezione di Entomologia,
MSNM) assisted us in getting micro-measurements with
the optical micrometer; Ermano Bianchi (Laboratorio di
Tassidermia, MSNM) dissected some alligator hatchlings
and birds for comparative anatomical observations; Ste-
fany Potze and Lemmy Mashinini (Department of Her-
petology, Transvaal Museum, Pretoria, South Africa) ar-
ranged the shipping of the Nile crocodile hatchling bones
collected by Fritz Huchzermeyer (Onderstepoort Veteri-
nary Institute).

Our very special, heartfelt thanks go to Marco Audi-
tore, who passionately drew all the anatomical drawings
of Scipionyx, improving them dozens of times according
to our continuous new interpretations. Marco’s boss, Ful-
vio Montaldo (Cantieri Navali di Sestri s.r.l., Genova),
patiently endured the days off that Marco took in order to
travel to the MSNM and study the specimen almost like
us. In addition to Marco, who also worked with Arianna
Nicora, many thanks are due to all other palaeoartists who
accepted our proposal to produce their artwork free of
charge, specifically for the present monograph: Davide
Bonadonna, Paolo Cinquemani, Fabio Fogliazza, Lukas
Panzarin, Fabio Pastori, Tullio Perentin, Loana Riboli,
Emiliano “Troco” and Renzo Zanetti.

For helpful discussions we are grateful to a number
of researchers (here listed in alphabetical order): Amy
Balanoff, Vivianne Bernardo de Sant’ Anna, Paulo Brito,
Chris Brochu, Mauro Buttafava, Matt Carrano, Andrea
Cau, Luis Chiappe, Dan Chure, Leon Claessens, Phil Cur-
rie, Massimo Delfino, Mike D’Emic, Cinzia Domeneghi-
ni, Greg Erickson, Alessandro Garassino, Stephen Ga-
tesy, Alan Gishlick, Ursula Gòhlich, Tom Holtz, Fritz
Huchzermeyer, John Hutchinson, Alex Kellner, Dave
Martill, Jesùs Marugan, Francesco Mascarello, Octavio
Matéus, Frangois Meunier, Umberto Nicosia, Gregory
Paul, Scott Persons, Fabio Petti, Karin Peyer, Federico
Pezzotta, Stefano Scali, Mary Schweitzer, Giorgio Teruz-
zi, David Vesely, Larry Witmer and Xu Xing.

Among others, Phil Currie (University of Alberta,
Edmonton) and Giorgio Teruzzi (MSNM) encouraged
us in the past years to carry on with the study of the
Pietraroja theropod; with the same aim, Stefania Nosotti
(MSNM) translated for us some literature published in
German.

Our bibliographic database improved with the help
of Marco Auditore, Derek Briggs, Andrea Cau, Fulvio
Gandolfi, Paola Livi, Lukas Panzarin, Mary Schweitzer
and many others; Giorgio Chiozzi, Nicolai Christiansen,
Lorenzo Crosta, and Luigi Scaccabarozzi provided use-
ful references; Karen Gariboldi and Daniela Cipollone
assisted us in arranging the paper-based, and the elec-
tronic archive.

Last but not least, we wish to thank also all persons
who helped us in various ways in the past years, apolo-
gising if we have overlooked anybody: Enza Braca, Ser-
gio Bravi, Cesare Brizio, Luigi Cagnolaro, Luigi Caz-
zaro, Maria Fariello, Fabio Fogliazza, Giovanna Gange-
mi, Henry Gee, Antonio Giannattasio, David Govoni,
Eva Koppelhus, Anastasios Kotsakis, Peter Makovicky,
Giuseppe Marramà, Rodolfo Martini, Ralph Molnar, Er-
manno Montanara, Ferdinando Moretti Foggia, Franco
Nodo, Mark Norell, Michael Novacek, Renato Pollini,
Antonella Russo, Gianni Tafuni, Silvo Tassinari and
Franco Valoti.

Special thanks

This monograph was written and published with
the financial support of the Società Italiana di Scienze
Naturali, the Museo di Storia Naturale di Milano and
the following additional sponsors, painstakingly made
aware of the importance of our work by our colleague
Ilaria Vinassa Guaraldi de Regny: Bocconi Traverso
& Partners; Beth Campbell Lasio, Giovanni Lasio and
their friends; Cinehollywood S.r.l.; Hasbro Ine., IMCD
Italia S.r.l.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 193

REFERENCES

AA.VV., 2005 — Nomina Anatomica Veterinaria. The Editorial
Committee.

Agnolin F. D., Ezcurra M. D., Pais D. F. & Salisbury S. W.,
2010 — A reappraisal of the Cretaceous non-avian dinosaur
faunas from Australia and New Zealand: evidence for their
Gondwanan affinities. Journal of Systematic Palaeontology,
8 (2): 257-300.

Andreassi G., Claps M., Sarti M., Nicosia U. & Venturo D.,
1999 — The late Cretaceous Dinosaur tracksite near Altamu-
ra (Bari), Southern Italy. Geoitalia 1999, 2° Forum Italiano
di Scienze della Terra. Riassunti, 1: 28.

Arduini P., 1993 — Research on Upper Permian Reptiles of Saka-
mena Formation (Madagascar). In: Evolution, ecology and
biogeography of the Triassic Reptiles. Mazin J. M. & Pinna
G. (eds.). Paleontologia Lombarda, N. Ser., II: 5-8.

Avanzini M., Leonardi G., Masetti D. & Mietto P., 2000 — Con-
clusioni. In: Dinosauri in Italia. Le orme giurassiche dei La-
vini di Marco (Trentino) e gli altri resti fossili italiani. Leo-
nardi G. & Mietto P. (eds.). Accademia Editoriale: 393-398.

Azuma Y. & Currie P. J., 2000 — A new carnosaur (Dinosauria:
Theropoda) from the Lower Cretaceous of Japan. Canadian
Journal of Earth Sciences, 37: 1735-1753.

Balanoff A. M. & Rowe T., 2007 — Osteological description of
an embryonic skeleton of the extinct elephant bird, Aepy-
ornis (Palaeognathae: Ratitae). Society of Vertebrate Pale-
ontology Memoir, 9: 1-53.

Balanoff A. M., Norell M. A., Grellet-Tinner G. & Lewin M.
R., 2008 — Digital preparation of a probable neoceratopsian
preserved within an egg, with comments on microstructural
anatomy of ornithischian eggshells. Naturwissenschaften,
95: 493-500.

Balanoff A. M., Xu X., Kobayashi Y., Matsufune Y. & Norell
M. A., 2009 — Cranial Osteology of the Theropod Dinosaur
Incisivosaurus gauthieri (Theropoda: Oviraptorosauria).
American Museum Novitates, 3651: 1-35.

Barbera C. & La Magna G., 1999 — Finding of ammonites in
the ittiolithic limestones at Pietraroja. In: Third International
Symposium on Lithographic Limestones. Renesto S. (ed.).
Rivista del Museo Civico di Scienze Naturali “E. Caffi”, 20:
23-24.

Barbera C. & Macuglia L., 1988 — Revisione dei tetrapodi del
Cretacico inferiore di Pietraroia (Matese orientale, Bene-
vento) appartenenti alla collezione Costa del Museo di Pale-
ontologia dell’Università di Napoli. Memorie della Società
Geologica Italiana, 41: 567-574.

Barbera C. & Macuglia L., 1991 — Cretaceous herpetofauna of
Pietraroia. In: Symposium on the Evolution of Terrestrial
Vertebrates. Ghiara G. (ed.). Unione Zoologica Italiana, Se-
lected Symposia and Monographs, 4: 421-429.

Barsbold R., 1983 — Carnivorous dinosaurs from the Cretaceous
of Mongolia. Trudy Sovmestnoi Sovetsko-Mongolskoi Pale-
ontologicheskoi Ekspeditsii, 19: 1-117. [In Russian].

Barsbold R. & Osmélska H., 1999 — The skull of Velociraptor
(Theropoda) from the Late Cretaceous of Mongolia. Acta
Paleontologica Polonica, 44: 189-219.

Barsbold R. & Perle A., 1984 — The first record of a primitive
ornithomimosaur from the Cretaceous of Mongolia. Paleon-
tological Journal, 1984: 118-120.

Barsbold R., Maryanska T. & Osmolska H., 1990 — Ovirapto-
rosauria. In: The Dinosauria. Weishampel D. B., Dodson
P. & Osmélska H. (eds.). University of California Press:
249-258.

Bartiromo A., Barone Lumaga M. R. & Bravi S., 2006 — First
finding of a fossil fern (Matoniaceae) in the paleontologi-

cal site of Pietraroja (Benevento, Southern Italy). Bollettino
della Società Paleontologica Italiana, 45 (1): 29-34.

Bassani F., 1885 — Risultati ottenuti dallo studio delle prin-
cipali ittiofaune cretaciche. Rendiconti dell'Istituto Lom-
bardo, 18: 513-535.

Baumel J. J. & Raikow R. J., 1993 — Arthrologia. In: Handbook
of Avian Anatomy: Nomina Anatomica Avium. Baumel J. J.,
King A. S., Breazile J. E., Evans H. E. & Vanden Berge J. C.
(eds.). Nuttal Ornithological Club: 133-187.

Baumel J. J., King A. S., Breazile J. E., Evans H. E. & Vanden
Berge J. C., 1993 — Handbook of Avian Anatomy: Nomina
Anatomica Avium. Nuttal Ornithological Club.

Bellairs A. & Jenkins C. R., 1960 — The skeleton of birds. In:
Biology and comparative physiology of birds. Marshall A.
J. (ed.). Academic Press: 241-300.

Bennett S. C., 1993 — The ontogeny of Preranodon and other
pterosaurs. Paleobiology, 19: 92-106.

Benson R. B. J., Carrano M. T. & Brusatte S. L., 2010 — A new
clade of archaic large-bodied predatory dinosaurs (Thero-
poda: Allosauroidea) that survived to the latest Mesozoic.
Naturwissenschaften, 97: 71-78.

Benton M. J., Cook E., Grigorescu D., Popa E. & Tallédi E.,
1997 — Dinosaurs and other tetrapods in an Early Cretaceous
bauxite-filled fissure, northwestern Romania. Palaeogeog-
raphy, Palaeoclimatology, Palaeoecology, 130: 275-292.

Berger A. J., 1960 — The musculature. In: Biology and com-
parative physiology of birds. Marshall A. J. (ed.). Academic
Press: 301-344.

Bever G. S. & Norell M. A., 2009 — The perinate skull of By-
ronosaurus (Troodontidae) with observations on the cranial
ontogeny of paravian theropods. American Museum Novi-
tates, 3657: 1-51.

Biur K. & Thapliyal J. P., 1972 — Cranial Pneumatization in the
Indian Weaver Bird, P/oceus philippinus. The Condor, 74
(2): 198-200.

Blamires S. J., 2004 — Habitat Preferences of Coastal Goan-
nas (Varanus panoptes): Are They Exploiters of Sea Turtle
Nests at Fog Bay, Australia? Copeia, 2: 370-377.

Bonaparte J. F., Novas F. E. & Coria R. A., 1990 — Carnotaurus
sastrei Bonaparte, the horned, lightly built carnosaur from the
Middle Cretaceous of Patagonia. Natural History Museum of
Los Angeles County, Contributions in Science, 416: 1-42.

Bosellini A., 2002 — Dinosaurs ‘re-write” the geodynamics of
the eastern Mediterranean and the paleogeography of the
Apulia Platform. Earth Science Review, 59: 211-234.

Botte V. & Pelagalli G. V., 1982 — Anatomia funzionale degli
uccelli domestici. Edizioni Ermes.

Bottjer D. J., Etter W., Hagadorn J. W., Tang C. M. (eds.), 2002 —
Exceptional fossil preservation. Columbia University Press.

Brainerd E. L., 1999 — New perspectives on the evolution of
lung ventilation mechanisms in vertebrates. Experimental
Biology Online, 4: 11-28.

Bravi S., 1987 — Contributo allo studio del giacimento ad ittioli-
ti di Pietraroja (Benevento). Dipartimento di Paleontologia,
Università degli Studi di Napoli “Federico II”, unpublished
degree thesis.

Bravi S., 1988 — Contributo allo studio del giacimento ad Ittio-
liti di Pietraroja (Benevento). I. P/europholis decastroi n.
sp. (Pisces, Actinopterygii, Pholidophoriformes). Memorie
della Società Geologica Italiana, 41 (1): 575-586.

Bravi S., 1994 — New observations on the Lower Cretaceous
fish Notagogus pentlandi Agassiz (Actinopterygii, Haleco-
stomi, Macrosemiidae). Bollettino della Società Paleonto-
logica Italiana, 33: 51-70.

194 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Bravi S., 1999 — A tentative reassessment of the fauna and flora
from the Pietraroja plattenkalk (Bn). In: Third International
Symposium on Lithographic Limestones. Renesto S. (ed.).
Rivista del Museo Civico di Scienze Naturali “E. Caf”, 20:
39-41.

Bravi S., & De Castro P., 1995 — The Cretaceous fossil fish-
es level of Capo d’Orlando, near Castellammare di Stabia
(NA). Biostratigraphy and depositional environment. Me-
morie di Scienze Geologiche, 47: 45-72.

Bravi S. & Garassino A., 1998 — New biostratigraphic and pale-
oecological observations on the Plattenkalk of the Lower
Cretaceous (Albian) of Pietraroia (Benevento, S. Italy), and
its decapod crustacean assemblage. Atti della Società Ita-
liana di Scienze Naturali e del Museo Civico di Storia Natu-
rale di Milano, 138/1997 (I-II): 119-171.

Breislak S., 1798 — Topografia fisica della Campania.Stamperia
Antonio Brazzini.

Briggs D. E. G., 2003 — The role of decay and mineralization
in the preservation of soft-bodied fossils. Annual Review of
Earth and Planetary Sciences, 31: 275-301.

Briggs D. E. G. & Kear A. J., 1993a— Fossilization of soft tissue
in the laboratory. Science, 259: 1439-1442.

Briggs D. E. G. & Kear A. J., 1993b — Decay and preservation
of polychaetes: taphonomic thresholds in soft-bodied organ-
isms. Paleobiology, 19: 107-135.

Briggs D. E. G. & Wilby P. R., 1996 — The role of the calcium
carbonate-calcium phosphate switch in the mineralization
of soft-bodied fossils. Journal of Geological Society, 153:
665-668.

Briggs D. E. G., Kear A. J., Martill D. M. & Wilby P. R., 1993 —
Phosphatization of soft tissue in experiments and fossils.
Journal of the Geological Society, 150: 1035-1038.

Briggs D. E. G., Wilby P. R., Pérez-Moreno B. P., Sanz J. L. &
Martinez M. F., 1997 — The mineralization of dinosaur soft
tissue in the Lower Cretaceous of Las Hoyas, Spain. Journal
of the Geological Society, 154 (4): 587-588.

Britt B. B., 1991 — Theropods of Dry Mesa Quarry (Morrison
Formation, Late Jurassic), Colorado, with emphasis on the
osteology of Torvosaurus tanneri. Brigham Young Univer-
sity Geology Studies, 37: 1-72.

Britt B. B. & Naylor B. G., 1994 — An embryonic Camarasau-
rus (Dinosauria, Sauropoda) from the Upper Jurassic Mor-
rison Formation (Dry Mesa Quarry, Colorado). In: Dinosaur
eggs and babies. Carpenter K.., Hirsch K.. F. & Horner J. R.
(eds.). Cambridge University Press: 256-264.

Brochu C. A., 1996 — Closure of neurocentral sutures during croc-
odilian ontogeny: implications for maturity assessment in fos-
sil archosaurs. Journal of Vertebrate Paleontology, 16: 49-62.

Brochu C. A., 2003 — Osteology of 7yrannosaurus rex: insights
from a nearly complete skeleton and high-resolution com-
puted tomographic analysis of the skull. Society of Verte-
brate Paleontology Memoir, 7: 1-138.

Bryant H. N. & Russell A. P., 1992 — The role of phylogenetic
analysis in the inference of unpreserved attributes of ex-
tinct taxa. Philosophical Transactions of the Royal Society,
Series B, 337: 405-418.

Brusatte S. L., Carr T. D., Erickson G. M., Bever G. S. &
Norell M. A., 2009 — A long-snouted, multihorned tyran-
nosaurid from the Late Cretaceous of Mongolia. Proceed-
ings of the National Academy of Sciences of the United
States of America, 106 (41): 17261-17266.

Burnham D. A., Derstler K. L., Currie P. J., Bakker R. T., Zhou
Z-H. & OstromJ. H., 2000 — Remarkable new birdlike dino-
saur (Theropoda: Maniraptora) from the Upper Cretaceous
of Montana. University of Kansas Paleontological Contri-
butions, New Series, 13: 1-14.

Butler J. R. & Upchurch P., 2007 — Highly incomplete taxa
and the phylogenetic relationships of the theropod dinosaur
Juravenator starki. Journal of Vertebrate Paleontology, 27
(1): 253-256.

Butterfield N. J., 2002 — Leanchoilia guts and the interpretation
of three-dimensional structures in Burgess Shale-type fos-
sils. Paleobiology, 28: 155-171.

Caldwell M. W. & Dal Sasso C., 2004 — Soft-tissue preservation
in a 95 million year old marine lizard: form, function, and
aquatic adaptation. Journal of Vertebrate Paleontology, 24
(4): 980-985.

Canudo J. I., Barco J. L., Pereda-Suberbiola X., Ruiz-Omefiaca
J. I, Salgado L., Fernandez-Baldor F. T. & Gasulla J. M.,
2009 — What Iberian dinosaurs reveal about the bridge said
to exist between Gondwana and Laurasia in the Early Cre-
taceous. Bulletin de la Société Géologique de France, 180
(I)

Carannante G., D’Argenio B., Dello Iacovo B., Ferreri V.,
Mindszenty A. & Simone L., 1988 — Studi sul carsismo
cretacico dell’ Appennino Campano. Memorie della Società
Geologica Ita-liana, 41: 733-759.

Carannante G., Pugliese A., Simone L. & Vigorito M., 2004 —
A Cretaceous tectonically controlled carbonate margin: the
case history of the Matese Mountains, central-southern Ap-
ennines, Italy. 23° AS Meeting of Sedimentology, Coimbra.
Abstract Volume: 79.

Carannante G., Signore M. & Vigorito M., 2006 — Vertebrate-
rich Plattenkalk of Pietraroia (Lower Cretaceous, Southern
Apennines, Italy): a new model. Facies, 52: 555-577.

Carpenter K., 1994 — Baby Dryosaurus from the Upper Jurassic
Morrison Formation of Dinosaur National Monument. In:
Dinosaur eggs and babies. Carpenter K., Hirsch K. F. &
Horner J. R. (eds.). Cambridge University Press: 288-297.

Carpenter K.., 1999 — Eggs, nests, and baby dinosaurs. A look at
dinosaur reproduction. Indiana University Press.

Carpenter K., 2002 — Forelimb biomechanics of nonavian thero-
pod dinosaurs in predation. Senckenbergiana lethaea, 82
(1): 59-76.

Carpenter K., Miles C., & Cloward K., 2005a — New small
theropods from the upper Jurassic Morrison Formation of
Wyoming. In: The Carnivorous Dinosaurs. Carpenter K.
(ed.). Indiana University Press: 23-48.

Carpenter K., Miles C., Ostrom J. H. & Cloward K., 2005b —
Redescription of the small maniraptoran theropods Orni-
tholestes and Coelurus from the Upper Jurassic Morrison
Formation of Wyoming. In: The Carnivorous Dinosaurs.
Carpenter K. (ed.). Indiana University Press: 49-71.

Carr T. D. & Williamson T. E., 2004 — Diversity of late Maas-
trichtian Tyrannosauridae (Dinosauria: Theropoda) from
western North America. Zoological Journal of the Linnean
Society, 142: 479-523.

Carrano M. T. & Hutchinson J. R., 2002 — Pelvic and Hindlimb
Musculature of 7yrannosaurus rex (Dinosauria: Theropoda).
Journal of Morphology, 253: 207-228.

Carrano M. T. & Sampson S. D., 2008 - The phylogeny of Cerato-
sauria. Journal of Systematic Palaeontology, 6 (2): 183-236.

Carrano M. T., Sampson S. D. & Forster C. A., 2002 — The oste-
ology of Masiakasaurus knopfleri, a small abelisauroid (Di-
nosauria: Theropoda) from the Late Cretaceous of Mada-
gascar. Journal of Vertebrate Paleontology, 22: 510-534.

Carrano M. T., Hutchinson J. R. & Sampson S. D., 2005 — New
information on Segisaurus halli, a small theropod dinosaur
from the Early Jurassic of Arizona. Journal of Vertebrate
Paleontology, 25: 835-849.

Carrier D. R. & Farmer C. G., 2000 — The evolution of pelvie
aspiration in archosaurs. Paleobiology, 26: 271-293.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 195

Carroll R. L., 1977 — The origin of lizards. In: Problems in ver-
tebrate evolution. Andrews S. M., Miles R. S. & Walker A.
D. (eds.). Linnean Society Symposium Series no. 4, Academ-
ic Press: 359-396.

Carroll R. L., 1988 — Vertebrate Paleontology and Evolution.
John Wiley & Sons.

Catenacci E. & Manfredini M., 1963 — Osservazioni strati-
grafiche sulla Civita di Pietraroia (Benevento). Bo//ettino
della Società Geologica Italiana, 82: 65-92.

Cati A., Sartorio D. & Venturini S., 1989 — Carbonate platforms
in the subsurface of the Northern Adriatic Area. Memorie
della Società geologica Italiana, 40: 295-308.

Cavitt L. C., Youm G. W., Meullenet J. F., Owens C. M. & Xiong
R., 2004 — Prediction of poultry meat tenderness using razor
blade shear, allo-Kramer shear, and sarcomere length. Jour-
nal of Food Science, 69 (1): 1-8.

Channel J. E. T., D’Argenio B. & Horwàth F., 1979 — Adria, the
African Promontory, in Mesozoic Mediterranean Paleoge-
ography. Earth Science Reviews, 15: 213-292.

Charig A. J. & Milner A. C., 1986 — Baryonyx, a remarkable
new theropod dinosaur. Nature, 324: 359-361.

Chen P. J., Dong Z. & Zhen S., 1998 — An exceptionally well-
preserved theropod dinosaur from the Yixian Formation of
China. Nature, 391: 147-152.

Chiappe L. M., Norell M. A. & Clark J. M., 2002 — The Creta-
ceous, short-armed Alvarezsauridae: Mononvykus and its kin.
In: Mesozoic birds: above the heads of dinosaurs. Chiappe
L.M. & Witmer L. M. (eds.). University of California Press:
87-119.

Chin K. & Bishop J., 2007 — Exploited twice: bored bone in a
theropod coprolite from the Jurassic Morrison Formation of
Utah, USA. In: Sediment-Organism Interactions: A_Multi-
faceted Ichnology. Bromley R. G., Buatois L. A., Mangano
M. G., Genise J. F. & Melchor R. N. (eds.). SEPM Special
Publications, 88: 379-387.

Chin K., Tokaryk T. T., Erickson G. M. & Calk L. C., 1998 -—A
king-sized theropod coprolite. Nature, 393: 680-682.

Chin K., Eberth D. A., Schweitzer M. H., Rando T. A., Sloboda
W. J. & Horner J. R., 2003 — Remarkable preservation of
undigested muscle-tissue within a late Cretaceous tyran-
nosaurid coprolite from Alberta, Canada. Palaios, 18 (3):
286-294.

Chinsamy A. & Hillenius W. J., 2004 — Physiology of nonavian
dinosaurs. In: The Dinosauria. 2°° edition. Weishampel D.
B., Dodson P. & Osmélska H. (eds.). University of Califor-
nia Press: 643-659.

Choiniere J. N., Clark J. M., Forster C. A. & Xu X., 2010-A
basal coelurosaur (Dinosauria: Theropoda) from the Late
Jurassic (Oxfordian) of the Shishugou Formation in Wu-
caiwan, People's Republic of China. Journal of Vertebrate
Paleontology, 30 (6): 1773-1796.

Chure D. J., 2001 — The wrist of A//osaurus (Saurischia: Thero-
poda), with observations on the carpus in theropods. In: New
perspectives on the origin and early evolution of birds: pro-
ceedings of the International Symposium in Honor of John
H. Ostrom. Gauthier J. & Gall L. F. (eds). Peabody Museum
of Natural History, Yale University: 283-300.

Chure D. J. & Madsen J. H., 1996 — On the presence of furculae
in some nonmaniraptoran theropods. Journal of Vertebrate
Paleontology, 16: 63-66.

Cillari A., Di Stefano P., Guzzetta D., Nicosia U., Petti FM. &
Zarcone G., 2009 — Back to Adria, the African Promontory:
geological and palaeontological constraints from central and
southern Italy. International Conference on Vertebrate Pal-
aeobiogeography and continental bridges across Tethys, Me-
sogea, and Mediterranean Sea. Museo Geologico Giovanni

Capellini, Dipartimento di Scienze della Terra e Geologico-
Ambientali (28-29 September 2009). Abstract Book: 28-30.

Claessens L. P. A. M., 2004 — Dinosaur gastralia; origin, mor-
phology, and function. Journal of Vertebrate Paleontology,
24 (1): 89-106.

Clark J. M., Perle A., & Norell M. A., 1994 — The skull of Er-
likosaurus andrewsi, a Late Cretaceous “segnosaur” (Thero-
poda: Therizinosauridae) from Mongolia. American Muse-
um Novitates, 3115: 1-39.

Clark J. M., Norell M. A. & Chiappe L., 1999 — An oviraptorid
skeleton from the Late Cretaceous of Ukhaa Tolgod, Mon-
golia, preserved in an avian-like brooding position over an
oviraptorid nest. American Museum Novitates, 3265: 1-36.

Clark J. M., Norell M. A. & Barsbold R., 2001 — Two new ovi-
raptorids (Theropoda: Oviraptorosauria) Late Cretaceous
Djadoktha Formation, Ukhaa Tolgod, Mongolia. Journal of
Vertebrate Paleontology, 21: 209-213.

Clark J. M., Norell M. A. & Rowe T., 2002 — Cranial Anatomy
of Citipati osmolskae (Theropoda, Oviraptorosauria), and a
reinterpretation of the holotype of Oviraptor philoceratops.
American Museum Novitates, 3364: 1-26.

Cocude-Michel M., 1963 — Les Rhynchocephales et les Sauriens
des Calcaires Lithographiques (Jurassique-Superieur) d’ Eu-
rope Occidentale. Nouvelles Archives du Museum d'Histoire
Naturelle de Lyon, T: 1-187.

Codd J. R., Manning P. L., Norell M. A. & Perry S. F., 2007 —
Avian-like breathing mechanics in maniraptoran dinosaurs.
Proceeding of the Royal Society B, 1233: 1-5.

Colbert E. H., 1989 — The Triassic dinosaur Coe/ophysis. Mu-
seum of Northern Arizona Bulletin, 57: 1-160.

Colbert E. H. & Russell D. A., 1969— The small Cretaceous dinosaur
Dromaeosaurus. American Museum Novitiates, 2380: 1-49.
Conti M. A., Morsilli M., Nicosia U., Sacchi E., Savino V.,
Wagensommer A., Di Maggio L. & Gianolla P., 2005 —
Jurassic Dinosaur Footprints From Southern Italy: Foot-
prints as Indicators of Constraints in Paleogeographic Inter-

pretation. Palaios, 20 (6): 534-550.

Coombs Jr. W. P., 1982 — Juvenile specimens of the ornithis-
chian dinosaurs Psittacosaurus. Palaeontology, 25: 89-107.

Coria R. A. & Currie P. J., 2006 — A new carcharodontosaurid
(Dinosauria, Theropoda) from the Upper Cretaceous of Ar-
gentina. Geodiversitas, 28 (1): 71-118.

Costa O. G., 1851 — Cenni intorno alle scoperte fatte nel Regno
riguardante la Paleontologia. // Filiatre Sebezio, 21: 40-46.

Costa O. G., 1853-1864 — Paleontologia del Regno di Napoli,
I-II. Atti Accademia Pontaniana, 5, 7, 8.

Costa O. G., 1865— Studi sopra i terreni ad Ittioliti delle Provincie
napolitanedirettiastabilirel’etàgeologicade’ medesimi. Parte
II: Calcarea stratosa di Pietraroja. Atti dell’Accademia delle
Scienze Fisiche e Matematiche di Napoli, s. 2, 2 (16): 1-33.

Costa O. G., 1866 — Nuove osservazioni e scoperte intor-
no ai fossili della calcarea ad ittioliti di Pietraroja. Atti
dell’Accademia delle Scienze Fisiche e Matematiche di Na-
pollis.192:(22) 112:

Cross H. R., West R. L. & Dutson T. R., 1981 — Comparison of
methods for measuring sarcomere length in beef semitendi-
nosus muscle. Meat Science, 5: 261-266.

Currie P. J., 1995 — New information on the anatomy and rela-
tionships of Dromaeosaurus albertensis (Dinosauria: Thero-
poda). Journal of Vertebrate Paleontology, 15: 576-591.

Currie P. J., 2003 — Cranial anatomy of tyrannosaurid dinosaurs
from the Late Cretaceous of Alberta, Canada. Acta Palaeon-
tologica Polonica, 48 (2): 191-226.

Currie P. J. & Chen P. J., 2001 — Anatomy of Sinosauropteryx
prima from Liaoning, northeastern China. Canadian Jour-
nal of Earth Sciences, 38: 1705-1727.

196 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Currie P. J. & Dong Z., 2001 — New information on Cretaceous
troodontids (Dinosauria, Theropoda) from the People’s Re-
public of China. Canadian Journal of Earth Sciences, 38:
1753-1766.

Currie P. J. & Russell D. A., 1988 — Osteology and relationships
of Chirostenotes pergracilis (Saurischia, Theropoda) from
the Judith River (Oldman) Formation of Alberta, Canada.
Canadian Journal of Earth Sciences, 25: 972-986.

Currie P. J. & Zhao X. J., 1993a — A new carnosaur (Dinosauria,
Theropoda) from the Jurassic of Xinjang, People”s Republic of
China. Canadian Journal of Earth Sciences, 30: 2037-2081.

Currie P. J. & Zhao X. J., 1993b — A new troodontid (Dinosau-
ria, Theropoda) braincase from the Dinosaur Park Forma-
tion (Campanian) of Alberta. Canadian Journal of Earth
Sciences, 30: 2231-2247.

Currie P. J., Hurum J. H. & Sabath K., 2003 — Skull structure
and evolution in tyrannosaurid dinosaurs. Acta Palaeonto-
logica Polonica, 48 (2): 227-234.

Dalla Vecchia F. M., 2001 — The Mesozoic Periadriatic Car-
bonate Platforms as terrestrial ecosystems: the vertebrate
evidence. VII International Symposium on Mesozoic Ter-
restrial Ecosystems, Buenos Aires, 26 September-1 October
1999, Abstracts: 20-21.

Dalla Vecchia F. M., 2002 — Cretaceous dinosaurs in the Adriat-
ic-Dinaric Carbonate Platform (Italy and Croatia): paleoen-
vironmental implications and paleogeographical hypotheses.
Memorie della Società Geologica Italiana, 57 (1): 89-100.

Dalla Vecchia F. M., 2005 — Between Gondwana and Laurasia:
Cretaceous Sauropods in an Intraoceanic Carbonate Plat-
form. In: Thunder Lizards: The Sauropodomorph Dino-
saurs. Carpenter K. & Tidwell V. (eds.). Indiana University
Press: 395-429.

Dalla Vecchia F. M., 2009 — Tethyshadros insularis, a new had-
rosauroid dinosaur (Ornithischia) from the Upper Creta-
ceous of Italy. Journal of Vertebrate Paleontology, 29 (4):
1100-1116.

Dalla Vecchia F. M., Tarlao A., Tunis G. & Venturini S., 2000
— New dinosaur track sites in the Albian (Early Cretaceous)
of the Istrian Peninsula (Croatia). Memorie di Scienze Geo-
logiche, 52: 193-292.

Dalla Vecchia F. M., Tarlao A., Tunis G. & Venturini S., 2001
— Dinosaur track sites in the upper Cenomanian (Late Cre-
taceous) of the Istrian peninsula (Croatia). Bollettino della
Società Paleontologica Italiana, 40: 25-54.

Dal Sasso C., 2001 — Dinosauri italiani. Marsilio Editore.

Dal Sasso C., 2002 — Update on Italian dinosaurs: witnesses
of a terrestrial, epicontinental ecosystem. 8% International
Symposium on Mesozoic Terrestrial Ecosystems. Cape Town
(South Africa), 21-26 July 2002, Abstract Book: 24.

Dal Sasso C., 2003 — Les dinosaures d’Italie. In: Les dinosaures
d’Europe/European Dinosaurs. Padian K., de Ricqlès A. &
Taquet T. (eds.). Comptes Rendus Palevol, Académie des
Sciences de Paris, 2 (1): 45-66.

Dal Sasso C., 2004 — Dinosaurs of Italy. Indiana University
Press.

Dal Sasso C. & Maganuco S., 2009 — Osteology, ontogenetic as-
sessment, phylogeny, paleobiology, and soft-tissue anatomy
of Scipionyx samniticus. Journal of Vertebrate Paleontol-
ogy, 29 (3, suppl.): 84A.

Dal Sasso C. & Signore M., 1998a — Exceptional soft-tissue
preservation in a theropod dinosaur from Italy. Nature, 392:
383-387.

Dal Sasso C. & Signore M., 1998b — Scipionyx samniticus
(Theropoda, Coelurosauria) and its exceptionally preserved
internal organs. Journal of Vertebrate Paleontology, 18 (3,
suppl.): 37A.

D’Argenio B., 1963 - I calcari ad ittioliti del Cretacico inferiore
del Matese. Atti dell’Accademia delle Scienze Fisiche e Ma-
tematiche, 4 (3): 1-63.

D’Argenio B., 1976 — Le piattaforme carbonatiche periadria-
tiche. Una rassegna di problemi nel quadro geodinamico
mesozoico dell’area mediterranea. Memorie della Società
Geologica Italiana, 13 (2, suppl.): 137-159.

D’Argenio B., Pescatore T. & Scandone P., 1973 — Schema ge-
ologico dell’ Appennino Meridionale (Campania, Lucania).
Accademia Nazionale dei Lincei, Quaderno, 183: 49-72.

D’Emic M., 2009 — The evolution of tooth replacement rates in
sauropod dinosaurs. Journal of Vertebrate Paleontology, 29
(3, suppl.): 84A.

D’Erasmo G., 1914 — La fauna e l’età dei calcari ad ittioliti di
Pietraroia (Prov. di Benevento). Palaeontographia italica,
20: 29-86.

D’Erasmo G., 1915 — La fauna e l’età dei calcari ad ittioliti di
Pietraroia (Prov. di Benevento). Palaeontographia italica,
2 15353:

de Klerk W. J., Forster C. A., Sampson S. D., Chinsamy A. &
Ross C. F., 2000 — A new coelurosaurian dinosaur from the
Early Cretaceous of South Africa. Journal of Vertebrate
Paleontology, 20: 324-332.

DeLamater E. D. & Courtenay W. R. Jr., 1974 — Fish scales as
seen by scanning electron microscopy. Florida Scientist, 37:
141-149.

Dercourt J., Ricou L. E. & Vrielynck B., 1993 — Atlas Tethys
Palaeoenvironmental Maps. P. Gauthier-Villars.

Dercourt J., Gaetani M., Vrielynck B., Barriere E., Biju-Duval
B., Brunet M. F., Cadet J. P., Crasquin S. & Sandulescu M.
E., 2000 — Atlas Peri-Tethys, Palaeogeographical Maps.
Carte Géologique du Monde / Commission for the Geologic
Map of the World.

de Ricglès A. J., Padian K., Horner J. R. & Francillon-Viellot
H., 2000 — Paleohistology of the bones of pterosaurs (Rep-
tilia: Archosauria): anatomy, ontogeny, and biomechanical
implications. Zoological Journal of the Linnean Society,
129: 349-385.

Diefenbach C., 1975 — Gastric function in Caiman crocodilus
(Crocodylia: Reptilia). I. Rate of gastric digestion and gas-
tric motility as a function of temperature. Comparative Bio-
chemistry and Physiology, SA: 259-265.

D’Orazi Porchetti S., Conti M. A., Nicosia U., Mariotti N., Petti
F. M., Sacchi E. & Valentini M., 2008 — Italian Vertebrate
Ichnology Reference List. Studi Trentini di Scienze Natu-
rali, Acta Geologica, 83: 335-347.

Dunham R. J., 1962 — Classification of carbonate rocks accord-
ing to depositional texture. In: Classification of carbonate
rocks. Ham W. E. (ed.). American Association of Petroleum
Geologists, Memoir, 1: 108-121.

Duncker H. R., 1972 — Structure of the avian lungs. Respiration
Physiology, 14 (1-2): 44-63.

Edmund A. G., 1969 — Dentition. In: Biology of the Reptilia,
Vol. 1: Morphology A. Gans C., Bellairs A. & Parson T. S.
(eds.). Academic Press: 117-200.

Elzanowski A., 2002 — Archaeopterygidae (Upper Jurassic of
Germany). In: Mesozoic birds: above the heads of dino-
saurs. Chiappe L. M. & Witmer L. M. (eds.). University of
California Press: 129-159.

Erickson G. M., 1996a — Daily deposition of dentine in juvenile
Alligator and assessment of tooth replacement rates using in-
cremental line counts. Journal of Morphology, 228: 189-194.

Erickson G. M., 1996b — Incremental lines of von Ebner in di-
nosaurs and the assessment of tooth replacement rates using
growth line counts. Proceedings of the National Academy of
Science, 93: 14623-14627.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 197

Erickson G. M., Curry-Rogers K. & Yerby S. A., 2001 — Dino-
saurian growth patterns and rapid avian growth rates. Na-
ture, 412: 429-433.

Estes R., 1983 — Sauria Terrestria, Amphisbaenia. In: Handbuch
der Palioherpetologie 10A. Wellnhofer P. (ed.). Gustav
Fischer Verlag.

Estes R., de-Queiroz K. & Gauthier J. A., 1988 — Phylogenetic
relationships within Squamata. In: Phylogenetic relation-
ships of the lizard families. Estes R. & Pregill G. (eds.).
Stanford University Press: 119-282.

Evans S. E., Raia P. & Barbera C., 2004 — New lizards and rhyn-
chocephalians from the Lower Cretaceous of southern Italy.
Acta Palaeontologica Polonica, 49 (3): 393-408.

Evans S. E., Raia P. & Barbera C., 2006 — The Lower Creta-
ceous lizard genus Chometokadmon from Italy. Cretaceous
Research, 27: 673-683.

Farmer C. G., 2006 — On the origin of avian air sacs. Respira-
tory Physiology & Neurobiology, 154: 89-106.

Farmer C. G. & Sanders K., 2010 — Unidirectional Airflow in
the Lungs of Alligators. Science, 327: 338-340.

Finetti I. R., 2005 — Ionian and Alpine Neotethyan Ocean open-
ing. In: CROP PROJECT: Deep Seismic Exploration of the
Central Mediterranean and Italy. Finetti I. R. (ed.). Elsevier:
103-107.

Fisher P. E., Russell D. A, Stoskopf M. K, Barrick R. E., Ham-
mer M. & Kuzmitz A. A., 2000 — Cardiovascular Evidence
for an Intermediate or Higher Metabolic Rate in an Ornithi-
schian Dinosaur. Science, 288: 503-505.

Freels D., 1975 — Plattenkalk-Becken bei Pietraroia (Prov. Be-
nevento, S-Italien) als Voraussetzung einer Fossillagerstàt-
ten-Bildung. Neues Jahrbuch fiir Geologie und Palciontolo-
gie, Abhandlungen, 148: 320-352.

Frey E., 1988 — Anatomie des Kérperstammes von A/ligator
mississippiensis Daudin. Stuttgarter Beitriige zur Naturkun-
de A, 24: 1-106.

Frey E., Tischlinger H., Buchy M.-C. & Martill D. M., 2003 —
New specimens of Pterosauria (Reptilia) with soft parts with
implications for pterosaurian anatomy and locomotion. In:
Evolution and Palaeobiology of Pterosaurs. Buffetaut E.
& Mazin J.-M. (eds.). Geological Society Special Publica-
tions, 217: 233-266.

Galton P. M., 1990 — Basal Sauropodomorpha-Prosauropoda.
In: The Dinosauria. Weishampel D. B., Dodson P. & Osmdl-
ska, H. (eds.). University of California Press: 320-344.

Garassino A. & Schweigert G., 2006 — The Upper Jurassic Soln-
hofen decapod crustacean fauna: review of the types from old
descriptions. Part I. Infraorders Astacidea, Thalassinidea, and
Palinura. Memorie della Società Italiana di Scienze Naturali e
del Museo Civico di Storia Naturale di Milano, 34 (1): 3-44.

Garilli V., Klein N., Buffetaut E., Sander P. M., Pollina F., Gal-
letti L., Cillari A. & Guzzetta D., 2009 — First dinosaur bone
from Sicily identified by histology and its paleobiogeo-
graphical implications. Neues Jahrbuch fiir Geologie und
Palciontologie, Abhandlungen, 252 (2): 207-216.

Gatesy S. M., 1990 — Caudofemoral musculature and the evolu-
tion of theropod locomotion. Paleobiology, 16: 170-186.
Gauthier J., 1986 — Saurischian monophyly and the origin of
birds. In: The Origin of Birds and the Evolution of Flight.
Padian K. (ed.). Memoirs of the Californian Academy of Sci-

ences, 8: 1-55.

Gauthier J. A., Estes R. & de-Queiroz K., 1988 — A phyloge-
netic analysis of Lepidosauromorpha. In: Phylogenetic rela-
tionships of the lizard families. Estes R. & Pregill G. (eds.).
Stanford University Press: 15-98.

Geist N. R. & Jones D. G., 1996 — Juvenile skeletal structure and
the reproductive habits of dinosaurs. Science, 272: 712-714.

George J. C. & Berger A. J., 1966 — Avian myology. Academic
Press.

Gianolla P., Morsilli M. & Bosellini A., 2000a — First discov-
ery of Early Cretaceous dinosaur footprints in the Gargano
Promontory (Apulia carbonate platform, southern Italy).
In: Quantitative Models on Cretaceous Carbonates and the
Eastern Margin of the Apulia Platform. Global Sedimentary
Geology Program (GSGP), Cretaceous Resources Events
and Rhytms (CRER) Working Group 4 (eds.). Vieste, Gar-
gano, Italy. September 25-28" 2000, Abstract Book: 9.

Gianolla P., Morsilli M., Dalla Vecchia F. M., Bosellini A. &
Russo A., 2000b — Impronte di dinosauri nel Cretaceo infe-
riore del Gargano (Puglia, Italia Meridionale): nuove impli-
cazioni paleogeografiche. 80° Riunione Estiva della Società
Geologica Italiana, Trieste, 6-8 settembre 2000. Riassunti:
265-266.

Gianolla P., Morsilli M. & Bosellini A., 2001 — Impronte di di-
nosauri nel Gargano. In: Cenni di Geologia e itinerari geo-
logici. Il Promontorio del Gargano. Bosellini A. & Morsilli
M. (eds.). Box 1.2.

Gishlick A. D. & Gauthier J. A., 2007 — On the manual mor-
phology of Compsognathus longipes and its bearing on the
diagnosis of Compsognathidae. Zoological Journal of the
Linnean Society, 149: 569-581.

Godfrey S. J. & Currie P. J., 2004 — A theropod (Dromaeosauri-
dae, Dinosauria) sternal plate from the Dinosaur Park Forma-
tion (Campanian, Upper Cretaceous) of Alberta, Canada. In:
Feathered Dragons. Currie P. J., Koppelhus E. B., Shugar M.
A. & Wright J. L. (eds.). Indiana University Press: 144-149.

Géhlich U. B. & Chiappe L. M., 2006 — A new carnivorous di-
nosaur from the Late Jurassic Solnhofen archipelago. Na-
ture, 440: 329-332.

Gonhlich U. B., Tischlinger H. & Chiappe L. M., 2006 — Jurave-
nator starki (Reptilia, Theropoda), ein neuer Raubdinosau-
rier aus dem Oberjura der sidlichen Frankenalb (Silddeut-
schland): Skelettanatomie und Weichteilbefunde. Archaeop-
teryx, 24: 1-26.

Goodwin M. B., Clemens W. A., Horner J. R. & Padian K.., 2006 —
The smallest known 7riceratops skull: new observations on
ceratopsid cranial anatomy and ontogeny. Journal of Verte-
brate Paleontology, 26: 103-112.

Gradstein F. M., Ogg J. G. & Smith A. G. (eds.), 2004 — A Geo-
logic Time Scale 2004. Cambridge University Press.

Grigorescu D., Venezel N., Csiki Z. & Limberea R., 1999 —
New latest Cretaceous microvertebrate fossil assemblages
from the Hateg basin (Romania). Geologie en Mijnbouw,
78: 310-314.

Hirschler A., Lucas J. & Hubert J.-C., 1990 — Bacterial involve-
ment in apatite genesis. FEMS Microbiology Ecology, 73:
211-220.

Hoffstetter R. & Gasc J. P., 1969 — Vertebrae and ribs of modern
reptiles. In: Biology of the Reptilia. Gans C. (ed.). Academic
Press, 1: 201-310.

Holliday C. M., 2009 — New insights into dinosaur jaw muscle
anatomy. Anatomical Record, 292: 1246-1265.

Holtz T. R. Jr., 1994 — The phylogenetic position of the Tyran-
nosauridae: implication for theropod systematics. Journal of
Paleontology, 68: 1100-1117.

Holtz T. R. Jr., 1996 — Phylogenetic taxonomy of the Coelu-
rosauria (Dinosauria: Theropoda). Journal of Paleontology,
70: 536-538.

Holtz T. R. Jr., 2000 — A new phylogeny of the carnivorous di-
nosaurs. Gaia, 15: 5-61.

Holtz T. R. Jr., 2004 — Tyrannosauroidea. In: The Dinosauria
(2° edition). Weishampel D. B., Dodson P. & Osmélska H.
(eds.). University of California Press: 111-136.

LI

198 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Holtz T. R. Jr., Molnar R. E. & Currie P. J., 2004 — Basal Tetanurae.
In: The Dinosauria (2"° edition). Weishampel D. B., Dodson P.
& Osmòlska H. (eds.). University of California Press: 71-110.

Hone D. W. E., Tischlinger H., Xu X. & Zhang F., 2010 — The
Extent of the Preserved Feathers on the Four-Winged Dino-
saur Microraptor gui under Ultraviolet Light. PLoS ONE, 5
(2)3e9223!

Horner J. R. & Currie P. J., 1994 — Embryonic and neonatal mor-
phology and ontogeny of a new species of Hypacrosaurus
(Ornithischia, Lambeosauridae) from Montana and Alberta.
In: Dinosaur eggs and babies. Carpenter K., Hirsch K. F. &
Horner J. R. (eds.). Cambridge University Press: 312-337.

Horner J. R., de Ricqlès A. & Padian K., 2000 — Long bone his-
tology of the hadrosaurid dinosaur Maiasaura peeblesorum:
growth dynamics and physiology based on an ontogenetic
series of skeletal elements. Journal of Vertebrate Paleontol-
ogy, 20 (1): 115-129.

Huchzermeyer F. W., 2000 — Le patologie dello struzzo e degli
altri ratiti. Papi Editore.

Huchzermeyer F. W., 2003 — Crocodiles: Biology, Husbandry
and Diseases. Centre for Agricultural Bioscience Interna-
tional Publishing.

Hutchinson J. R., 2001 — The evolution of femoral osteology
and soft tissues on the line to extant birds (Neornithes). Zoo-
logical Journal of the Linnean Society, 131: 169-197.

Hutchinson J. R. & Gatesy S. M., 2000 — Adductors, abductors,
and the evolution of archosaur locomotion. Paleobiology,
26: 734-751.

Hutt S., Martill D. A. & Barker M. J., 1996 — The first European
allosaurid dinosaur (Lower Cretaceous, Wealden Group,
England). Neues Jahrbuch fiir Geologie und Paldontologie,
Monatshefte, 1996 (10): 635-644.

Hwang S. H., Norell M. A., Qiang J. & Keqin G., 2002 — New
specimens of Microraptor zhaoianus (Theropoda: Dromae-
osauridae) from northeastern China. American Museum No-
vitates, 3381: 1-44.

Hwang S. H., Norell M. A., Qiang J. & Keqin G. A., 2004 —
Large compsognathid from the Early Cretaceous Yixian
Formation of China. Journal of Systematic Paleontology, 2
(1): 13-30.

Iniesto M., Lopez-Archilla A., Buscalioni A. D., Penalver E.,
Fregenal-Martìnez M. A. & Guerrero M. C., 2009 — Experi-
mental simulation of initial stages of fossilisation by bac-
terial sealing. Implications for the formation of Las Hoyas
Konservat-Lagerstitte (Lower Cretaceous, Iberian Ranges,
Spain). In: 5° International Symposium on Lithographic
Limestone and Plattenkalk. Billon-Bruyat J. P., Marty D.,
Costeur L., Meyer C.A. & Thuring B. (eds.). Actes 2009 bis
de la Société Jurassienne d’émulation: 46-47.

Jamroz D., Wertelecki T., Wiliczkiewicz A., Orda J., & Sko-
rupinska J., 2004 — Dynamics of yolk sac resorption and
post-hatching development of the gastrointestinal tract in
chickens, ducks and geese. Journal of Animal Physiology
and Animal Nutrition, 88 (5-6): 239-250.

Janes D. & Gutzke W. H. N., 2002 — Factors Affecting Reten-
tion Time of Turtle Scutes in Stomachs of American Alliga-
tors, A/ligator mississippiensis. American Midland Natural-
ist, 148 (1): 115-119.

Ji Q., Currie P. J., Norell M. A. & Ji S. A., 1998 — Two feathered
dinosaurs from North-eastern China. Nature, 393: 753-761.

Ji Q., Norell M. A., Makovicky P. J., Gao K., Ji S. & Yuan C., 2003 —
An early ostrich dinosaur and implication for ornithomimosaur
phylogeny. American Museum Novitates, 3420: 1-19.

Ji Q., Ji S., Lù J. & You H., 2005 — First avialian bird from
China, Jinfengopteryx elegans gen. et sp. nov. Geological
Bulletin of China, 24 (3): 197-210.

Ji Q., Ji S., Li J. & Yuan C., 2007a — A New Giant Compsognathid
Dinosaur with Long Filamentous Integuments from Lower Cre-
taceous of Northeastern China. Acta Geologica Sinica, 81: 8-15.

Ji S., Gao C., Liu J., Meng Q. & Ji Q., 2007b — New Material of
Sinosauropteryx (Theropoda: Compsognathidae) from West-
ern Liaoning, China. Acta Geologica Sinica, 81: 177-182.

Jianu C.-M. & Weishampel D. B., 1999 — The smallest of the
largest. A new look at possible dwarfing in sauropod dino-
saurs. Geologie en Mijnbou, 78: 335-343.

Kadler K. E., Holmes D. F., Trotter J. A. & Chapman J. A., 1996 —
Collagen fibril formation. Biochemical Journal, 316: 1-11.

Kardong K.V., 1997 — Vertebrates: Comparative Anatomy,
Function, Evolution. McGraw-Hill.

Kellner A. W. A., 1996a — Remarks on Brazilian dinosaurs.
Memoirs of the Queensland Museum, 39: 611-626.

Kellner A. W. A., 1996b — Fossilised theropod soft-tissue. Na-
ture, 379: 32.

Kellner A. W. A., 1999 — Short note on a new dinosaur (Thero-
poda, Coelurosauria) from the Santana Formation (Romual-
do Member, Albian), Northeastern Brazil. Boletim do Museu
Nacional, N.S., 49: 1-8.

Killops S. D. & Killops V. J., 1993 — An Introduction to Organic
Geochemistry. Longman Scientific and Technical.

Kirkland J. I. & Wolfe D. G., 2001 — First definitive therizino-
saurid (Dinosauria: Theropoda) from North America. Jour-
nal of Vertebrate Paleontology, 21: 410-414.

Kirkland J. I., Britt B. B., Whittle C. H., Madsen S. K. & Burge
D. L., 1998-A small coelurosaurian theropod from the Yel-
low Cat Member of the Cedar Mountain Formation (Lower
Cretaceous, Barremian) of eastern Utah. In: Lower and Mid-
dle Cretaceous Ecosystems. Lucas S. G., Kirkland J. I. &
Estep J. W. (eds.). New Mexico Museum of Natural History
and Science Bulletin, 14: 239 -248.

Kirkland J. I., Zanno L. E., Sampson S. D., Clark J. M. & De-
Blieux D. D., 2005 — A primitive therizinosauroid dinosaur
from the Early Cretaceous of Utah. Nature, 7038: 84-87.

Kitching I. J., Forey P. L., Humphries C. J., & Williams D. M.,
1998 — Cladistics. Second Edition. The Theory and Practice
of Parsimony Analysis. Oxford University Press: 1-228.

Klasing K. C., 1998 — Comparative avian nutrition. Centre for
Agricultural Bioscience International Publishing.

Kobayashi Y. & Barsbold R., 2005 — Reexamination of a primi-
tive ornithomimosaur, Garudimimus brevipes Barsbold, 1981
(Dinosauria:Theropoda), from the Late Cretaceous of Mon-
golia. Canadian Journal of Earth Sciences, 42: 1501-1521.

Kobayashi Y. & Li J.-C., 2003 — A new ornithomimid dinosaur
with gregarious habits from the Late Cretaceous of China.
Acta Palaeontologica Polonica, 48 (2): 235-259.

Kuchel L. J. & Franklin C. E., 2000 — Morphology ofthe cloaca in
the estuarine crocodile, Crocodylus porosus, and its plastic re-
sponse to salinity. Journal of Morphology, 245 (2): 168-176.

Kundràt M., Cruickshank A. R. I, Manning T. W. & Nudds J.,
2008 — Embryos of therizinosauroid theropods from the Up-
per Cretaceous of China: diagnosis and analysis of ossifica-
tion patterns. Acta Zoologica, 88: 231-251.

Lambe L. M., 1917 — The Cretaceous carnivorous dinosaur Gor-
gosaurus. Geological Survey of Canada, Memoir, 100: 1-84.

Langer M. C., 2004 — Basal Saurischia. In: The Dinosauria
(2" edition). Weishampel D. B., Dodson P. & Osmélska H.
(eds.). University of California Press: 25-46.

Larson P. & Rigby J. K. Jr., 2005 — Furcula of 7yrannosaurus
rex. In: The Carnivorous Dinosaurs. Carpenter K. (ed.). /n-
diana University Press: 247-255.

Leahy G., 2000 — Noses, Lungs, and Guts. In: The Scientific
American Book of Dinosaurs. Paul G. (ed.). Byron Preiss
Visual Publications & Scientific American: 52-63.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 199

Leonardi G., 2008 — Vertebrate ichnology in Italy. Studi Trentini
di Scienze Naturali, Acta Geologica, 83: 213-221.

Leonardi G. & Avanzini M., 1994 — Dinosauri in Italia. Le Sci-
enze (Quaderni), 76: 69-81.

Leonardi G. & Lanzinger M., 1992 — Dinosauri nel Trentino:
venticinque piste fossili nel Liassico di Rovereto (Trento,
Italia). Paleocronache, 1: 13-24.

Leonardi G. & Mietto P., 2000 — Le piste liassiche di dino-
sauri dei Lavini di Marco. In: Dinosauri in Italia. Le orme
giurassiche dei Lavini di Marco (Trentino) e gli altri resti
fossili italiani. Leonardi G. & Mietto P. (eds.). Accademia
Editoriale: 169-247.

Leonardi G. & Teruzzi G., 1993 — Prima segnalazione di uno
scheletro fossile di dinosauro (Theropoda, Coelurosauria) in
Italia (Cretacico di Pietraroia, Benevento). Pal/eocronache,
1: 7-14.

Li D., Norell M. A., Gao K.-Q.., Smith N. D. & Makovicky P. J.,
2009 — A longirostrine tyrannosauroid from the Early Creta-
ceous of China. Proceedings of the Royal Society of London,
B, 277: 183-190.

Lindgren J., Currie P. J., Siverson M., Rees J., Cederstròm P. &
Lindgren F., 2007 — The first neoceratopsian dinosaur re-
mains from Europe. Palaeontology, 50 (4): 929-937.

Lingham-Soliar T. & Wesley-Smith J., 2008 — First investigation
of the collagen D-band ultrastructure in fossilized vertebrate
integument. Proceedings of the Royal Society of London, B,
275: 2207-2212.

Lipkin C., Sereno P. C. & Horner J. R., 2007 — The furcula in
Suchomimus tenerensis and Tyrannosaurus rex (Dinosauria:
Theropoda: Tetanurae). Journal of Paleontology, 81 (6):
1523-1527.

Long, J. A. & McNamara K. J., 1997- Heterochrony. In: Ency-
clopedia of dinosaurs. Currie P. J. & Padian K. (eds.). Aca-
demic Press: 311-317.

Longrich N. R. & Currie P. J., 2009 — A/bertonvkus borealis,
a new alvarezsaur (Dinosauria: Theropoda) from the Early
Maastrichtian of Alberta, Canada: Implications for the sys-
tematics and ecology of the Alvarezsauridae. Cretaceous
Research, 30: 239-252.

Lyman R. L., 1994 — Vertebrate Taphonomy. Cambridge Uni-
versity Press.

Madsen J. R. Jr., 1976 — Allosaurus fragilis: a revised osteology.
Utah Geological Survey Bulletin, 109: 1-163.

Madsen J. H. Jr. & Welles S. P., 2000 — Ceratosaurus (Dinosau-
ria, Theropoda): a revised osteology. Miscellaneous Publi-
cations of the Utah Geological Survey, 2: 1-80.

Maganuco S., Cau A. & Pasini G., 2005 — First description of
theropod remains from the Middle Jurassic (Bathonian) of
Madagascar. Atti della Società Italiana di Scienze Naturali
e del Museo Civico di Storia Naturale in Milano, 146 (11):
165-202.

Maganuco S., Steyer J. S., Pasini G., Boulay M., Lorrain S.,
Bénéteau A. & Auditore M., 2009 — An exquisite specimen
of Edingerella madagascariensis (Temnospondyli) from the
Lower Triassic of NW Madagascar: cranial anatomy, phyl-
ogeny and restorations. Memorie della Società Italiana di
Scienze Naturali e del Museo Civico di Storia Naturale di
Milano, XXXVI (II): 1-72.

Maina J. N., 2005 — The lung-air sac system of birds: develop-
ment, structure, and function. Springer-Verlag.

Makovicky P. J., 1995 — Phylogenetic aspects of the vertebral
morphology of Coelurosauria (Dinosauria: Theropoda).
Master s Thesis, Copenhagen University.

Makovicky P. J., 1997 — Postcranial axial skeleton, compara-
tive anatomy. In: Encyclopedia of dinosaurs. Currie P. J. &
Padian K. (eds.). Academic Press: 579-590.

Makovicky P. J. & Currie P. J., 1998 — The presence of a furcula
in tyrannosaurid theropods, and its phylogenetic and func-
tional implications. Journal of Vertebrate Paleontology, 18:
143-149.

Makovicky P. J. & Norell M. A., 1998 — A partial ornithomimid
braincase from Ukhaa Tolgod (Upper Cretaceous, Mongo-
lia). American Museum Novitates, 3247: 1-16.

Makovicky P. J. & Norell M. A., 2004 — Troodontidae. In: The
Dinosauria (2"° edition). Weishampel D. B., Dosdson P. & Os-
mòlska H., (eds.). University of California Press: 184-195.

Makovicky P. J., Norell M. A., Clark J. M. & Rowe T., 2003 —
Osteology and Relationships of Byronosaurus jaffei (Thero-
poda: Troodontidae). American Museum Novitates, 3402:
1-32.

Makovicky P. J., Kobayashi Y. & Currie P. J., 2004 — Ornithomi-
mosauria. In: The Dinosauria (2° edition). Weishampel D.
B., Dodson P. & Osmélska H. (eds.). University of Califor-
nia Press: 137-150.

Manning P. L., 2008 — Grave secrets of dinosaurs. Soft tissues
and hard science. National Geographic Books.

Marsh O. C., 1881 — Jurassic birds and their allies. American
Journal of Science, 3" ser., 22: 337-340.

Martill D. M., 1988 — Preservation of fish in the Cretaceous
Santana formation of Brazil. Pal/aeontology, 31: 1-18.

Martill D. M., Frey E., Sues H-D. & Cruickshank A. R. I., 2000 —
Skeletal remains of a small theropod dinosaur with asso-
ciated soft structures from the Lower Cretaceous Santana
Formation of northeast Brazil. Canadian Journal of Earth
Sciences, 37 (6): 891-900.

Martin R. E., 1999 — Taphonomy: a process approach. Cam-
bridge University Press.

Martinez R. D. & Novas F. E., 2006 — Aniksosaurus darwini gen.
et sp. nov., a new coelurosaurian theropod from the Early Late
Cretaceous of Central Patagonia, Argentina. Revista del Mu-
seo Argentino de Ciencias Naturales, n.s. 8 (2): 243-259.

Mateus O., 1998 — Lourinhanosaurus antunesi, a new Upper Ju-
rassic Allosauroid (Dinosauria: Theropoda) from Lourinhà,
Portugal. Memdòrias de Academia de Ciéncias das Lisboa,
37:111-124.

Mateus I., Mateus H., Antunes M. T., Mateus O., Taquet P., Ri-
beiro V. & Manuppella G., 1998 — Upper Jurassic theropod
dinosaur embryos from Lourinhà (Portugal). Memorias de
Academia de Ciéncias das Lisboa, 37: 101-110.

McGowan G. J., 2002 — Albanerpetontid amphibians from the
Lower Cretaceous of Spain and Italy: a description and re-
consideration of their systematics. Zoological Journal of the
Linnean Society, 135: 1-32.

McGowan G. & Evans S. E., 1995 — Albanerpetontid amphib-
ians from the Cretaceous of Spain. Nature, 373: 143-145.

McLelland J., 1989 — Anatomy of lungs and air sacs. In: Form
and Function in Birds. Vol. 4. King A. S. & McLelland J.
(eds.). Academic Press: 221-279.

McLelland. J., 1990 — A colour atlas of avian anatomy. Wolfe
Publishing.

McNab B. K.., 2002 — The Physiological Ecology of Vertebrates:
A View from Energetics. Cornell! University Press.

Meunier F., 1984 — Structure et minéralisation des écailles de
quelques Osteoglossidae (Osteichthiens, Téléostéens). An-
nales des Sciences Naturelles, Zoologie, 3 Sér., 6: 111-124.

Mezga A. & Bajraktarevié Z., 1999 — Cenomanian dinosaur
tracks on the islet of Fenoliga in southern Istria, Croatia.
Cretaceous Research, 20: 735-746.

Mezga A., Meyer C. A., Cvetko Tesovié B., Bajraktarevié Z.
& Gusié I., 2006 — The first record of dinosaurs in the Dal-
matian part (Croatia) of the Adriatic-Dinaric carbonate plat-
form (ADCP). Cretaceous Research, 27: 735-742.

200 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Mezga A., Cvetko Tesovié B. & Bajraktarevié Z., 2007 — First
record of dinosaurs in the late Jurassic of the Adriatic-Dinar-
ic Carbonate Platform (Croatia). Palaios, 22 (2): 188-199.

Mietto P., 1988 — Piste di dinosauri nella Dolomia Principale
(Triassico superiore) del Monte Pelmetto (Cadore). Memo-
rie della Società Geologica Italiana, 30: 307-310.

Mindszenty A., D’Argenio B. & Aiello G., 1995 — Lithospheric
bulges recorded by regional unconformities. The case of
Mesozoic-Tertiary Apulia. Tectonophysics, 252: 137-161.

Mortimer M., 2004-2010 — Theropod Database. http://home.
comcast.net/-eoraptor/Home.html

Muhl Z. F., Grimm A. F. & Glick P., 1976 — Technique for meas-
urements of sarcomere length of striated muscle. Journal of
Dental Research, 55 (1): 170.

Murphy N. L., Trexler D. & Thompson M., 2007 — “Leonardo”,
amummified Brachylophosaurus (Ornithischia: Hadrosauri-
dae) from the Judith River Formation of Montana. In: Horns
& Beaks. Ceratopsian and Ornithopod Dinosaurs. Carpenter
K. (ed.). Indiana University Press: 117-133.

Murray A. M., Simons E. L. & Attiac Y. S., 2005 — A new clu-
peid fish (Clupeomorpha) from the Oligocene of Fayum,
Egypt, with notes on some other fossil clupeomorphs. Jour-
nal of Vertebrate Paleontology, 25 (2): 300-308.

Naish D. W., Hutt S., & Martill D. M., 2001 — Saurischian dino-
saurs: Theropods. In: Dinosaurs of the Isle of Wight. Field
Guides to Fossils 10. Martill D. M. & Naish D. W. (eds.).
Palaeontological Association: 242-309.

Naish D., Martill D. M. & Frey E., 2004 — Ecology, systematics
and biogeographical relationships of dinosaurs, including a
new theropod, from the Santana Formation (?Albian, Early
Cretaceous) of Brazil. Historical Biology, 16: 57-70.

Nesbitt S. J., Turner A. H., Spaulding M., Conrad J. L. & Norell
M. A., 2009 — The theropod furcula. Journal of Morphol-
ogy, 270 (7): 856-579.

Nicholls E. L. & Russell A. P., 1981 — A new specimen of
Struthiomimus altus from Alberta, with comments on the
classificatory characters of Upper Cretaceous ornithomim-
ids. Canadian Journal of Earth Sciences, 18: 518-526.

Nicosia U., Marino M., Mariotti N., Muraro C., Panigutti S., Pet-
ti F. M. & Sacchi E., 2000a — The Late Cretaceous dinosaur
tracksite near Altamura (Bari, southern Italy). I Geological
framework. Geologica Romana, 35 (1999): 231-236.

Nicosia U., Marino M., Mariotti N., Muraro C., Panigutti S.,
Petti F. M. & Sacchi E., 2000b — The Late Cretaceous dino-
saur tracksite near Altamura (Bari, southern Italy). II Apu-
losauripus federicianus new ichnogen. and new ichnosp.
Geologica Romana, 35 (1999): 237-247.

Nicosia U., Avanzini M., Barbera C., Conti M. A., Dalla Vecchia
F., Dal Sasso C., Gianolla P., Leonardi G., Loi M., Mariotti
N., Mietto P., Morsilli M., Paganoni A., Petti F. M., Piubelli
D., Raia P., Renesto S., Sacchi E., Santi G. & Signore M.,
2005 - I vertebrati continentali del Paleozoico e Mesozoico.
In: Paleontologia dei vertebrati in Italia. Evoluzione biolo-
gica, significato ambientale e paleogeografia. Bonfiglio L.
(ed.). Memorie del Museo Civico di Storia Naturale di Vero-
na, 2° serie, sezione Scienze della Terra, 6: 41-66.

Nicosia U., Petti F. M., Perugini G., D’Orazi Porchetti S., Sac-
chi E., Conti M. A., Mariotti N. & Zarattini A., 2007 — Dino-
saur Tracks as Paleogeographic Constraints: New Scenarios
for the Cretaceous Geography of the Periadriatic Region.
Ichnos, 14: 69-90.

Nopesa F., 1903 — Neues uber Compsognathus. Neues Jahrbuch
fiir Geologie und Palciontologie, 16: 476-494.

Norell M. A. & Makovicky P. J., 1997 — Important features of
the dromaeosaur skeleton: information from a new speci-
men. American Museum Novitates, 3215: 1-28.

Norell M. A. & Makovicky P. J., 1999 — Important features of
the dromaeosaur skeleton, II. Information from newly col-
lected specimens of Velociraptor mongoliensis. American
Museum Novitates, 3282: 1-45.

Norell M. A. & Makovicky P. J., 2004 — Dromaeosauridae. In:
The Dinosauria (2° edition). Weishampel D. B., Dodson P. &
Osmolska H. (eds.). University of California Press: 196-209.

Norell M. A., Clark J. M., Dashzeveg D., Barsbold R., Chiappe
L. M., Davidson A. R., McKenna M. C., Perle A. & No-
vacek M. J., 1994 — A theropod dinosaur embryo and the
affinities of the Flaming Cliffs dinosaur eggs. Science, 266:
779-782.

Norell M. A., Clark J. M., Chiappe L. M. & Dashzeveg D., 1995 —
A nesting dinosaur. Nature, 378: 774-776.

Norell M. A., Makovicky P. J. & Clark J. M., 1997 — A Ve-
lociraptor wishbone. Nature, 389: 447.

Norell M. A, Makovicky P. J, & Clark J. M., 2000 — A new troo-
dontid theropod from Ukhaa Tolgod, Mongolia. Journal of
Vertebrate Paleontology, 20: 7-11.

Norell M. A., Clark J. M. & Chiappe L. M., 2001 — An em-
bryonic oviraptorid (Dinosauria: Theropoda) from the Up-
per Cretaceous of Mongolia. American Museum Novitates,
3315: 1-17.

Norell M. A., Makovicky P. J., Bever G. S., Balanoff A. M., Clark
J. M., Barsbold R., & Rowe T., 2009 — A Review of the Mon-
golian Cretaceous Dinosaur Saurornithoides (Troodontidae:
Theropoda). American Museum Novitates, 3654: 1-63.

Noto C., 2009 — The potential utility of authigenic minerals on
modern and fossil bones for environmental and taphonomic
analysis. Journal of Vertebrate Paleontology, 29 (3, suppl.):
156A.

Novas F. E., Ezcurra M. D. & Lecuona A., 2008a — Orkoraptor
burkei nov. gen. et sp., a large theropod from the Maastrich-
tian Pari Aike Formation, Southern Patagonia, Argentina.
Cretaceous Research, 29: 468-480.

Novas F. E., Pol D., Canale J. I., Porfiri J. D. & Calvo J. O.,
2008b — A bizarre Cretaceous theropod dinosaur from Pat-
agonia and the evolution of Gondwanan dromaeosaurids.
Proceedings of the Royal Society B, 276 (1659): 1101-1107.

O’Connor P. M., 2006 — Posteranial pneumaticity: an evaluation
of soft-tissue influences on the postceranial skeleton and the
reconstruction of pulmonary anatomy in archosaurs. Jowur-
nal of Morphology, 267: 1199-1226.

O’Connor P. M., 2007 — The postcranial axial skeleton of Ma-
jungasaurus crenatissimus (Theropoda: Abelisauridae) from
the Late Cretaceous of Madagascar. Society of Vertebrate
Paleontology Memoir, 8: 127-162.

O’Connor P. M. & Claessens L. P., 2005 — Basic avian pul-
monary design and flow-through ventilation in non-avian
theropod dinosaurs. Nature, 436: 253-256.

Organ C. L., 2006 — Biomechanics of ossified tendons in orni-
thopod dinosaurs. Pa/eobiology, 32 (4): 652-665.

Osborn H. F., 1916 — Skeletal adaptations of Ornitholestes,
Struthiomimus, Tyrannosaurus. Bulletin of the American
Museum of Natural History, 35: 733-771.

Osi A., Butler R. J. & Weishampel D. B., 2010 — A Late Creta-
ceous ceratopsian dinosaur from Europe with Asian affini-
ties. Nature, 456: 466-468.

Osmblska H., Maryanska T. & Barsbold R., 1972 — A new dino-
saur, Gallimimus bullatus n. gen., n. sp. (Ornithomimidae)
from the Upper Cretaceous of Mongolia. Palaeontologia
Polonica, 27: 103-143.

Osmolska H., Currie P. J. & Barsbold R., 2004 — Oviraptoro-
sauria. In: The Dinosauria (2"° edition). Weishampel D. B.,
Dodson P. & Osmélska H. (eds.). University of California
Press: 165-183.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 201

Ostrom J. H., 1969 — Osteology of Deinonychus anthyrropus,
an unusual theropod from the Lower Cretaceous of Mon-
tana. Bulletin of the Peabody Museum of Natural History,
30: 1-165.

Ostrom J. H., 1970 — Archaeopteryx: Notice of a “New” Speci-
men. Science, 170: 537-538.

Ostrom J. H., 1978 — The osteology of Compsognathus longipes
Wagner. Zitteliana, 4: 73-118.

Padian K., 2004 — Basal Avialae. In: The Dinosauria (2°° edi-
tion). Weishampel D. B., Dodson P. & Osmblska H. (eds.).
University of California Press: 210-231.

Padian K., de Ricqlès A. J. & Horner J. R., 2001 — Dinosaurian
growth rates and bird origins. Nature, 412: 405-408.

Page R. D. M., 1996 - TREEVIEW: An application to display
phylogenetic trees on personal computers. Computer Appli-
cation in the Biosciences, 12: 357-358.

Page R. D. M., 2001 —- NEXUS Data Editor. 0.5.0. http://tax-
onomy.zoology.gla.ac.uk/rod/NDE/nde.html.

Panchangam A., Claflin D. R., Palmer M. L. & Faulkner J. A.,
2008 — Magnitude of Sarcomere Extension Correlates with
Initial Sarcomere Length during Lengthening of Activated
Single Fibers from Soleus Muscle of Rats. Biophysical
Journal, 95 (4): 1890-1901.

Patacca E. & Scandone P., 2004 — A geological transect across
the Southern Apennines along the seismic line CROP 04. In:
Field Trip Guide Books. Post Congress P20, 32" IGC Flor-
ence, 20-28 August 2004. Guerrieri L., Rischia I. & Serva L.
(eds.). Memorie Descrittive della Carta Geologica d'Italia,
63 (4): 14-36.

Patacca E. & Scandone P., 2007 — Geology of the Southern Ap-
ennines. Bollettino della Società Geologica Italiana, Volume
speciale, 7: 75-119.

Paul G. S., 1988 — Predatory Dinosaurs of the World: A Com-
plete Illustrated Guide. Simon & Schuster: 1-464.

Paul G. S., 2001 — Were the respiratory complexes of predatory
dinosaurs like crocodilians or birds? In: New Perspectives
on the Origin and Early Evolution of Birds. Proceedings of
the International Symposium in Honor of John H. Ostrom.
Gauthier J. A. & Gall L. F. (eds.). Peabody Museum of Natu-
ral History: 463-482.

Paul G. S., 2002 — Dinosaurs of the air: the evolution and loss of
flight in dinosaurs and birds. The John Hopkins University
Press: 1-460.

Pérez-Moreno B. P., Sanz J. L., Buscalioni A. D., Moratalla J.
J., Ortega F. & Rasskin-Gutman D., 1994 — A unique multi-
toothed ornithomimosaur dinosaur from the Lower Creta-
ceous of Spain. Nature, 370: 363-367.

Perry S. F., 1998 — Lungs: comparative anatomy, functional mor-
phology, and evolution. In: Biology of the Reptilia, Volume 19,
Morphology G, Visceral Organs. Gans C. & Gaunt A. S. (eds.).
Society for the Study of Amphibians and Reptiles: 1-92.

Perry S. F., 2001 — Functional morphology of the reptilian and
avian respiratory system and its implications for theropod
dinosaurs. In: New Perspectives on the Origin and Early
Evolution of Birds. Proceedings of the International Sym-
posium in Honor of John H. Ostrom. Gauthier J. A. & Gall
L. F. (eds.). Peabody Museum of Natural History: 429-441.

Persons W. S. & Currie P. J., 2011 - The tail of 7yrannosaurus:
reassessing the size and locomotive importance of the M.
caudofemoralis in non-avian theropods. Anatomical Record,
294: 119-131.

Petti F. M., Conti M. A., D’Orazi Porchetti S., Morsilli M., Ni-
cosia U. & Gianolla P., 2008a — A theropod dominated ich-
nocoenosis from late Hauterivian-early Barremian of Borgo
Celano (Gargano Promontory, Apulia, southern Italy). Rivi-
sta Italiana di Paleontologia e Stratigrafia, 14 (1): 3-17.

Petti F. M., D’Orazi Porchetti S., Conti M. A., Nicosia U., Peru-
gini G. & Sacchi E., 2008b — Theropod and sauropod foot-
prints in the Early Cretaceous (Aptian) Apenninic Carbonate
Platform (Esperia, Lazio, Central Italy): a further constraint
on the palaeogeography of the Central Mediterranean area.
Studi Trentini di Scienze Naturali, Acta Geologica, 83: 323-
334.

Peyer K., 2004 — A reevaluation of the French Compsognathus
of the Tithonian of southeastern France and its phylogenetic
relationship with other compsognathids and coelurosaurs.
PhD Thesis, Muséum national d'Histoire naturelle, Paris.

Peyer K., 2006 — A reconsideration of Compsognathus from the
Upper Tithonian of Canjuers, Southeastern France. Journal
of Vertebrate Paleontology, 26 (4): 879-896.

Pinto e Silva C. J. R., Martins M. R. F. B. & Guazzelli Filho
J., 2008 — Study on cranial and caudal mesenteric arteries
in opossum (Didelphis albiventris). International Journal of
Morphology, 26 (3): 635-637.

Pol D. & Norell M.A., 2004 — A new gobiosuchid crocodyliform
taxon from the Cretaceous of Mongolia. American Museum
Novitates, 3458: 1-31.

Pol D. & Powell J. E., 2007 — Skull anatomy of Mussaurus
patagonicus (Dinosauria: Sauropodomorpha) from the Late
Triassic of Patagonia. Historical Biology, 19 (1): 125-144.

Poyato-Ariza F. J. & Wenz S., 2002 — A new insight into pycno-
dontiform fishes. Geodiversitas, 24 (1): 139-248.

Raath M. A., 1985 — The theropod Syntarsus and its bearing on
the origin of birds. In: The beginnings of birds: Proceedings
of the International Archaeopteryx Conference, Eichstàtt
1984. Hecht M. K., Ostrom J. H., Viohl G. & WelInhofer P.
(eds.). Freunde des Jura- Museums: 219-227.

Rauhut O. W. M., 2003 — The interrelationships and evolution
of basal theropod dinosaurs. Special Papers in Palaeontol-
ogy, 69: 1-213.

Rauhut O. W. M. & Fechner R., 2005 — Early development of
the facial region in a non-avian theropod dinosaur. Proceed-
ings of the Royal Society of London B, 272: 1179-1183.

Rauhut O. W. M. & Xu X., 2005 — The small theropod dinosaurs
Tugulusaurus and Phaedrolosaurus from the Early Creta-
ceous of Xinjiang, China. Journal of Vertebrate Paleontol-
ogy, 25: 107-118.

Rauhut O. W. M., Milner A. C. & Moore-Fay S., 2009 — Cranial
osteology and phylogenetic position of the theropod dino-
saur Proceratosaurus bradlevi (Woodward, 1910) from the
Middle Jurassic of England. Zoological Journal of the Lin-
nean Society: 1-41.

Reisz R. R., Scott D., Sues H.-D., Evans D. C. & Raath M. A.,
2005 — Embryos of an early Jurassic prosauropod dinosaur
and their evolutionary significance. Science, 309: 761-764.

Richardson K., Webb G. & Manolis C., 2000 — Crocodiles: In-
side and Out. Surrey Beatty and Sons.

Rieppel O., 1992a — Studies on skeleton formation in reptiles.
I. The post-embryonic development of the skeleton in Cyr-
todactylus pubisulcus (Reptilia: Gekkonidae). Journal of
Zoology, 227: 87-100.

Rieppel O., 1992b — Studies on skeleton formation in reptiles.
III. Patterns of ossification in the skeleton of Lacerta vivipara
Jacquin (Reptilia, Squamata). Fieldiana: Zoology, 68: 1-25.

Rieppel O., 1993 — Studies on skeleton formation in reptiles: pat-
terns of ossification in the skeleton of Chelydra serpentina
(Reptilia, Testudines). Journal of Zoology, 231: 487-509.

Rinehart L. F., Lucas S. G. & Hunt A. P., 2007 — Furculae in the
Late Triassic theropod dinosaur Coe/ophysis bauri. Palcion-
tologische Zeitschrift, 81 (2): 174-180.

RomerA. S., 1966 — Vertebrate paleontology. University of Chi-
cago Press.

202 CRISTIANO DAL SASSO & SIMONE MAGANUCO

RomerA. S. & Parsons T. S., 1977 — The Vertebrate Body. Holt
Saunders International.

Rosenbaum G., Lister G. & Duboz C., 2004 — The Mesozoic
and Cenozoic motion of Adria (central Mediterranean): a
review of constraints and limitations. Geodinamica Acta, 17
(225-139;

Rowe T., 1989 — A new species of the theropod dinosaur Syntar-
sus from the Early Jurassic Kayenta Formation of Arizona.
Journal of Vertebrate Paleontology, 9 (2): 125-136.

Rowe T., McBride E. F. & Sereno P. C., 2001 — Technical com-
ment: dinosaur with a heart of stone. Science, 291 (5505):
183:

Ruben J. A., Jones T. D., Geist N. R. & Hillenius W. J., 1997 —
Lung structure and ventilation in theropod dinosaurs and
early birds. Science, 278: 1267-1270.

Ruben J. A., Jones T. D., Geist N. R. & Hillenius W. J., 1998 —
Letter in response to criticisms of Ruben et al., 1997. Sci-
ence, 281: 47-48.

Ruben J. A., Dal Sasso C., Geist N. R., Hillenius W. J., Jones T.
D. & Signore M., 1999 — Pulmonary function and metabolic
physiology of theropod dinosaurs. Science, 283: 514-516.

Ruben J. A., Jones T. D. & Geist N. R., 2003 — Respiratory and
reproductive paleophysiology of dinosaurs and early birds.
Physiological and Biochemical Zoology, 76: 141-164.

Russell D. A. & Dong Z., 1993 — A nearly complete skeleton of
a new troodontid dinosaur from the Early Cretaceous of the
Ordo Basin, Inner Mongolia, People’s Republic of China.
Canadian Journal of Earth Sciences, 30: 2163-2173.

Sacchi E., Conti M. A., D’Orazi Porchetti S., Nicosia U., Pe-
rugini G., Petti F. M. & Logoluso A., 2006 — A new diverse
dinosaur footprint association from the Calcare di Bari
(Lower Cretaceous, Apulia, southern Italy): implication for
paleoecology and paleogeography. Giornate di Paleontolo-
gia 2006. Abstracts: 78.

Sacchi E., Conti M. A., D’Orazi Porchetti S., Logoluso A., Ni-
cosia U., Perugini G. & Petti F. M., 2009 — Aptian dinosaur
footprints from the Apulian platform (Bisceglie, Southern
Italy) in the framework of periadriatic ichnosites. Palaeoge-
ography, Palaeoclimatology, Palaeoecology, 271: 104-116.

Sagemann J., Bale S. J., Briggs D. E. G. & Parkes R. J., 1999 —
Controls on the formation of authigenic minerals in asso-
ciation with decaying organic matter: an experimental ap-
proach. Geochimica et Cosmochimica Acta, 63: 1083-1095.

Saladin K. S., 2010 — Anatomy & Physiology: A Unity of Form
and Function (5° Edition). McGraw-Hill.

Salgado L., Coria R. A., & Chiappe L. M., 2005 — Osteology of
the sauropod embryos from the Upper Cretaceous of Pat-
agonia. Acta Palaeontologica Polonica, 50 (1): 79-92.

Sampson S. D. & Witmer L. M., 2007 — Craniofacial anatomy
of Majungasaurus crenatissimus (Theropoda: Abelisauri-
dae) from the Late Cretaceous of Madagascar. Society of
Vertebrate Palaeontology, Memoir, 8: 32-102.

Sander M. P., Mateus O., Laven T. & Knétschke N., 2006 —
Bone histology indicates insular dwarfism in a new Late
Jurassic sauropod dinosaur. Nature, 441: 739-741.

Sanz J. L., Chiappe L. M., Perez-Moreno B. P., Moratalla J.
J., Hernandéz Carrasquilla F., Buscalioni A. D., Ortega F.,
Poyato-Ariza F., Rasskin-Gutman D. & Martinéz-Delclòs
X., 1997 — A nestling bird from the Lower Cretaceous of
Spain: implications for avian skull and neck evolution. Sci-
ence, 276: 1543-1546.

Schachner E. R., Lyson T. R. & Dodson P., 2009 — Evolution of
the Respiratory System in Nonavian Theropods: Evidence
from Rib and Vertebral Morphology. In: Unearthing the
Anatomy of Dinosaurs: New Insights into their Functional
Morphology and Paleobiology. Dodson P. (ed.). The Ana-

tomical Record: Advances in Integrative Anatomy and Evo-
lutionary Biology, Special Issue, 292 (9): 1501-1513.

Schaffner F., 1998 — The liver. In: Biology of Reptilia. Gans C.
& Gaunt A. S. (eds.). Society for the Study of Amphibians
and Reptiles Press, 19: 297-374.

Scheid P. & Piiper J., 1989 — Respiratory mechanics and air flow
in birds. In: Form and function in birds. Vol. 4. King A. S. &
McLelland J. (eds.). Academic Press: 369-391.

Schettino A. & Scotese C., 2002 — Global kinematic constraints
to the tectonic history of the Mediterranean region and sur-
rounding areas during the Jurassic and Cretaceous. In: Re-
construction of the evolution of the Alpine-Himalayan Oro-
gen. Rosenbaum G. & Lister G. S. (eds.). Journal of the
Virtual Explorer, 8: 149-168.

Schwarz D., Ikejiri T., Breithaupt B. H., Sander P. M. & Klein
N., 2007a — A nearly complete skeleton of an early juvenile
diplodocid (Dinosauria: Sauropoda) from the lower Mor-
rison Formation (Late Jurassic) of north central Wyoming
and its implications for early ontogeny and pneumaticity in
sauropods. Historical Biology, 19: 225-253.

Schwarz D., Wings O. & Meyer C. A., 2007b — Super sizing
the giants: first cartilage preservation at a sauropod dinosaur
limb joint. Journal of the Geological Society, 164: 61-65.

Schwarz D., Frey E. & Meyer C. A, 2007c — Pneumaticity and
soft-tissue reconstructions in the neck of diplodocid and di-
craeosaurid sauropods. Acta Palaeontologica Polonica, 52
(1): 167-188.

Schwarz-Wings D., Frey E. & Martin T., 2009 — Reconstruction
of the bracing system of the trunk and tail in Hyposaurine
Dyrosaurids (Crocodylomorpha; Mesoeucrocodylia). Jour
nal of Vertebrate Paleontology, 29 (2): 453-472.

Schweitzer M. H., Watt J. A., Avci R., Forster C. A., Krause D.
W., Knapp L., Rogers R. R., Beech I. & Marshall M., 1999 —
Keratin immunoreactivity in the Late Cretaceous bird Raho-
navis ostromi. Journal of Vertebrate Paleontology, 10 (4):
1122722:

Schweitzer M. H., Wittmeyer J. L. & Horner J. R., 2007 — Soft
tissue and cellular preservation in vertebrate skeletal ele-
ments from the Cretaceous to the present. Proceeding of the
Royal Society B, 274: 183-197.

Schweitzer M. H., Avci R., Collier T. & Goodwin M. B., 2008 —
Microscopic, chemical and molecular methods for examin-
ing fossil preservation. In: Paléogénétique en Paléontologie,
Archéologie et Paléoanthropologie: Contributions et Limi-
tes. Vigne J.-D. & Darlu P. (eds.). Comptes Rendus Palevol,
7 (2-3): 159-184.

Seilacher A., 1970 — Begriff und Bedeutung der Fossil-Lager-
stitten. Neues Jahrbuch Geol. Palcdiontol. Abh., 1970: 34-
39,

Senter P., 2007 — A new look at the phylogeny of the Coeluro-
sauria (Dinosauria: Theropoda). Journal of Systematic Pa-
laeontology, 1-35.

Sereno P. C., 2001 — Alvarezsaurids: birds or ornithomimo-
saurs? In: New Perspectives on the Origin and Early Evolu-
tion of Birds: Proceedings of the International Symposium
in Honor of John H. Ostrom. Gauthier J. & Gall L. F. (eds.).
Peabody Museum of Natural History: 69-98.

Sereno P. C. & Arcucci A. B., 1994 — Dinosaurian precursors
from the Middle Triassic of Argentina: Marasuchus lilloen-
sis, gen. nov. Journal of Vertebrate Paleontology, 14: 53-73.

Sereno P. C., Forster C. A., Rogers R. R., & Monetta A. M.,
1993 — Primitive dinosaur skeleton from Argentina and the
early evolution of Dinosauria. Nature, 361: 64-66.

Sereno P. C., Wilson J. A., Larsson H. C. E., Dutheil D. B. &
Sues H. D., 1994 — Early Cretaceous dinosaurs from the Sa-
hara. Science, 265: 267-271.

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 203

Sereno P. C., Martinez R. N., Wilson J. A., Varricchio D. J., Al-
cober O. A. & Larsson H. C. E., 2008 — Evidence for avian
intrathoracic air sacs in a new predatory dinosaur from Ar-
gentina. PLoS ONE, 3 (9), e3303.

Sereno P. C., Tan L., Brusatte S. L., Kriegstain H. J., Zhao X.
& Cloward K., 2009 — Tyrannosaurid Skeletal Design First
Evolved at Small Body Size. Sciencexpress, 17 September
2009.

Serventy D. L., Nicholls C. A. & Farner D. S., 1966 — Pneu-
matization of the cranium of the zebra finch Taeniopygia
castanotis. Ibis, 190 (4): 570-578.

Signore M., 1995 — Il teropode del Plattenkalk della Civita di
Pietraroia (Cretaceo inferiore, BN). Dipartimento di Pale-
ontologia, Università degli Studi di Napoli “Federico II”,
unpublished degree thesis.

Signore M., 2004 — Sample excavations in Pietraroja (lower
Cretaceous, Southern Italy) in 2001 and notes on the Pie-
traroja paleoenvironment. Pa/Arch, 2: 13-22.

Signore M., Barbera C., DeVita S. & La Magna G., 2001 — Tetra-
pod Fauna of Pietraroja Plattenkalk (Benevento, Southern
Italy). 6 European Workshop on Vertebrate Paleontology.
Florence and Montevarchi (Italy), September 19-22, 2001.
Abstract book: 53.

Signore M., Bucci E., Pede C. & Barbera C., 2005 — A new
ichthyodectid fish from the Lower Cretaceous of Pietraroja
(Southern Italy). PalArch, 5: 25-29.

Signore M., Pede C., Bucci E. & Barbera C., 2006 — First re-
port of the genus C/adocyclus in the Lower Cretaceous of
Pietraroja (Southern Italy). Bollettino della Società Paleon-
tologica Italiana, 45 (1): 141-146.

Skoczylas R., 1978 — Physiology of the digestive tract. In: Bi-
ology of the Reptilia, Vol 8: Physiology B. Gans C. (ed.).
Academic Press: 589-717.

Smith J. W., 1968 — Molecular pattern in native collagen. Na-
ture, 219: 157-158.

Smith J. B. & Dodson P., 2003 — A proposal for a standard ter-
minology of anatomical notation and orientation in fossil
vertebrate dentitions. Journal of Vertebrate Paleontology,
DS(1)31=12:

Stampfli G. M., 2005 — Plate tectonics of the Apulia-Adria Mi-
crocontinents. In: CROP PROJECT: Deep Seismic Explo-
ration of the Central Mediterranean and Italy. Finetti I. R.
(ed.). E/sevier: 747-776.

Stein K., Csiki Z., Curry-Rogers K., Weishampel D. B., Redel-
storffe R., Carballido J. L. & Sander P. M., 2010 — Small
body size and extreme cortical bone remodeling indicate
phyletic dwarfism in Magyarosaurus dacus (Sauropoda:
Titanosauria). Proceedings of the National Academy of Sci-
ences.

Steyer J. S., 2003 — A revision of the Early Triassic “Capito-
saurs” (Stegocephali, Stereospondyli) from Madagascar,
with remarks on their comparative ontogeny. Journal of
Vertebrate Paleontology, 23: 544-555.

Sues H.-D., 1977 — The skull of Velociraptor mongoliensis, a
small Cretaceous theropod dinosaur from Mongolia. Pald-
ontologische Zeitschrift, 51 (3/4): 173-184.

Swofford D. L., 2002 — PAUP*. Phylogenetic Analysis Using
Parsimony (*and other methods). Version 4. Sinauer Asso-
ciates.

Taylor M. P., Wedel M. J. & Naish D., 2009 — Head and
neck posture in sauropod dinosaurs inferred from extant
animals. Acta Palaeontologica Polonica, 54 (2): 213-
220.

Tegelaar E. W., de Leeuw J. W., Derenne S. & Largeau C., 1989 —
A reappraisal of kerogen formation. Geochimica et Cosmo-
chimica Acta, 53: 3103-3106.

o)

Tischlinger H., 2002 — Der Eichstàtter Archaeopteryx im lang-
welligen UV-Licht. [The Eichstàtt specimen of Archaeop-
teryx under longwave ultraviolet light]. Archaeopteryx, 20:
21-38.

Turco E., Schettino A., Nicosia U., Santantonio M., Di Stefano
P., Iannace A., Cannata D., Conti M. A., Deiana G., D’Orazi
Porchetti S., Felici F., Liotta D., Mariotti N., Milia A., Petti
F. M,, Pierantoni P. P., Sacchi E., Sbrescia V., Tommasetti
K., Valentini M., Zamparelli V. & Zarcone G., 2007 — Meso-
zoic Paleogeography of the Central Mediterranean Region.
Geoitalia 2007, VI Forum Italiano di Scienze della Terra,
Epitome, 2: 108.

Tykoski R. S., 2005 — Anatomy, Ontogeny, and Phylogeny of
Coelophysoid Theropods. PAD Thesis, University of Texas,
Austin.

Tykoski R. S. & Rowe T., 2004 — Ceratosauria. In: The Dino-
sauria (2"° edition). Weishampel D. B., Dodson P., & Osmdl-
ska H. (eds.). University of California Press: 47-70.

Tykoski R. S., Forster C. A., Rowe T., Sampson S. D., & Munyi-
hwa D., 2002 — A furcula in the coelophysoid theropod Syn-
tarsus. Journal of Vertebrate Paleontology, 22: 728-733.

Varricchio D. J., 1997 — Growth and embryology. In: Encyclo-
pedia of dinosaurs. Currie P. J. & Padian K. (eds.). Academic
Press: 282-288.

Varricchio D. J., 2001 — Gut contents from a Cretaceous ty-
rannosaurid: implications for theropod dinosaur digestive
tracts. Journal of Paleontology, 75: 401-406.

Varricchio D. J., Horner J. R. & Jackson F. D., 2002 — Embryos
and eggs for the Cretaceous theropod dinosaur 7roodon for-
mosus. Journal of Vertebrate Paleontology, 22: 564-576.

Vigorito M., Simone L. & Carannante G., 2003 — Tectoni-
cally controlled carbonate channelized slope complexes: a
Cretaceous-Miocene case-history from Matese Mountains
(Central-Southern Apennines, Italy). S/ope 2003-Submarine
Slope Systems: Processes, Products and Prediction. April
28-29, 2003, Liverpool, Abstracts: 95.

Webb G. J. W., Hollis G. J. & Manolis S. C., 1991 — Feeding,
growth, and food conversion rates of wild juvenile saltwater
crocodiles (Crocodylus porosus). Journal of Herpetology,
25: 462-473.

Wedel M. J., 2009 — Evidence for bird-like air sacs in sauris-
chian dinosaurs. Journal of Experimental Zoology, 311A
(8): 611-628.

Weishampel D. B., Dodson P. & Osmoélska H., 2004 — Intro-
duction. In: The Dinosauria (2"° edition). Weishampel D. B..,
Dodson P. & Osmolska H. (eds.). University of California
Press: 1-3.

Welles S. P., 1984 — Dil/ophosaurus wetherilli (Dinosauria,
Theropoda). Osteology and comparisons. Palaeontographi-
ca Abt. A, 185: 85-180.

Wellnhofer P., 1985 — Remarks on the digit and pubis problems
of Archaeopteryx. In: The beginnings of birds: proceedings
of the International Archaeopteryx Conference, Eichstàtt
1984. Hecht M. K., Ostrom J. H., Viohl G. & WelInhofer P.
(eds.). Freunde des Jura- Museums: 113-122.

Westergaard B. & Ferguson M. W. J., 1990 — Development of
dentition in A/ligator mississippiensis: upper jaw dental and
craniofacial development in embryos, hatchlings and young
juveniles, with a comparison to lower jaw development. The
American Journal of Anatomy, 187: 393-421.

Wheeler T. L. & Koohmaraie M., 1994 — Prerigor and postrigor
changes in tenderness of ovine longissimus muscle. Journal
of Animal Science, 72: 1232-1238.

Wilby P. R. & Briggs D. E. G., 1997 — Taxonomic trends in the
resolution of detail preserved in fossil soft tissues. Geobios,
20: 493-502.

204 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Wilkinson M., 2001a - TAXEQ3: software and documentation.
Department of Zoology, The Natural History Museum.

Wilkinson M., 2001b - REDCON 3.0: software and documenta-
tion. Department of Zoology, The Natural History Museum.

Wilson J. A., 1999 — A nomenclature for vertebral laminae in
sauropods and other saurischian dinosaurs. Journal of Ver-
tebrate Paleontology, 19: 639-653.

Witmer L. M., 1995 — The extant phylogenetic bracket and the
importance of reconstructing soft tissues in fossils. In: Func-
tional morphology in vertebrate paleontology. Thomason J.
J. (ed.). Cambridge University Press: 19-33.

Witmer L. M., 1997 — The evolution of the antorbital cavity in
archosaurs: a study in soft-tissue reconstruction in the fos-
sil record with an analysis of the function of pneumaticity.
Journal of Vertebrate Paleontology, Memoir 3: 1-73.

Woodward A. S., 1910 — On a skull of Megalosaurus from the
Great Oolite of Minchinhempton (Gloucestershire). Quar-
terly Journal of the Geological Society, 66: 111-115.

Xu X., 2006 — Feathered dinosaurs from China and the evolution
of major avian characters. Integrative Zoology, 1: 4-11.

Xu X., 2010 — Horned dinosaurs venture abroad. Nature, 465:
431-432.

Xu X. & Norell M. A., 2004 — A new troodontid dinosaur from
China, with avian-like sleeping posture. Nature, 431: 838-
841.

Xu X. & Norell M. A., 2006 — Non-Avian dinosaur fossils from
the Lower Cretaceous Jehol Group of western Liaoning,
China. Geological Journal, 41: 419-417.

Xu X. & Wu X.-C., 2001 — Cranial morphology of Sinorni-
thosaurus millenii Xu et al. 1999 (Dinosauria: Theropoda:
Dromaeosauridae) from the Yixian Formation of Liaon-
ing, China. Canadian Journal of Earth Sciences, 38: 1939-
1/52

Xu X., Zhou Z. & Wang X., 2000 — The smallest known non-
avian theropod dinosaur. Nature, 408: 705-708.

Xu X., Wang X.-L. & You H.-L., 2001 — A juvenile ankylosaur
from China. Naturwissenschaften, 88: 297-300.

Xu X., Cheng Y. N., Wang X. L. & Chang C. H., 2002a — An
unusual oviraptorosaurian dinosaur from China. Nature,
419: 291-293.

Xu X., Norell M. A., Wang X. L., Makovicky P.J. & Wu X. C.,
2002b — A basal troodontid from the Early Cretaceous of
China. Nature, 415: 780-784.

Xu X., Norell M. A., Kuang X., Wang X., Zhao Q. & Jia C.,
2004 — Basal tyrannosauroids from China and evidence for
protofeathers in tyrannosauroids. Nature, 431: 680-684.

Xu X., Clark J. M., Forster C. A., Norell M. A., Erickson G.
M., Eberth D. A., Jia C. & Zhao Q., 2006 — A basal tyran-
nosauroid from the Late Jurassic of China. Nature, 439:
715-718.

Xu X., Wang K. B., Zhao K. J. & Li D. J., 2010 — First ceratop-
sid dinosaur from China and its biogeographical interpreta-
tion. Chinese Science Bulletin, 55: 1631-1635.

Yates A. M. & Warren A. A., 2000 — The phylogeny of the
“higher” temnospondyls (Vertebrata: Choanata) and its im-
plications for the monophyly and origins of Stereospondyli.
Zoological Journal of the Linnean Society, 128: 77-121.

Yilmaz P. O., Norton I. O., Leary D. & Chuchla J., 1996 — Tec-
tonic evolution and paleogeography of Europe. In: Peri-
Tethys Memoir 2: Structure and Prospects of Alpine Basins
and Forelands. Ziegler P. A. & Horvàth F. (eds.). Memoires
du Muséum d'’Histoire Naturelle, 170: 47-60.

Zanno L. E., 2006 — The pectoral girdle and forelimb of the
primitive therizinosauroid Fa/carius utahensis (Thero-
poda, Maniraptora): analyzing evolutionary trends within
Therizinosauroidea. Journal of Vertebrate Paleontology,
26: 636-650.

Zanno L. E., 2010 — Osteology of Fal/carius utahensis (Dino-
sauria: Theropoda): characterizing the anatomy of basal
therizinosaurs. Zoological Journal of the Linnean Society,
158: 196-230.

Zappaterra E., 1990 — Carbonate paleogeographic sequences of
the periadriatic region. Bollettino della Società Geologica
Italiana, 109: 5-20.

Zappaterra E., 1994 — Source rock distribution model of the
Periadriatic Region. American Association of Petroleum
Geologists Bulletin, 78: 333-354.

Zarcone G. & Di Stefano P., 2008 — Mesozoic discontinuities
in the Panormide Carbonate Platform: constraints on the
palaeogeography of the central Mediterranean. Rendiconti
Online della Società Geologica Italiana, 2: 191-194.

Zhang F., Kearns S. L., Orr P. J., Benton M. J., Zhou Z., John-
son D., Xu X. & Wang X., 2010 — Fossilized melanosomes
and the colour of Cretaceous dinosaurs and birds. Nature,
463: 1075-1078.

Zhang X.-H., Xu X., Zhao X.-J., Sereno P. C., Kuang X.-W.
& Tan L., 2001 — A long-necked therizinosauroid dinosaur
from the Upper Cretaceous Iren Dabasu Formation of Nei
Mongol, People’s Republic of China. Vertebrata PalAsi-
atica, 10: 282-290.

Zheng X., Xu X., You H., Zhao Q. & Dong Z., 2009 — A short-
armed dromaeosaurid from the Jehol Group of China with
implications for early dromaeosaurid evolution. Proceed-
ings of the Royal Society B, 277: 211-217.

Zhou Z.-H., Wang X.-L., Zhang F.-C. & Xu X., 2000 — Important
features of Caudipteryx - evidence from two nearly complete
new specimens. Vertebrata PalAsiatica, 10: 241-254.

Zhou Z., Barrett P. M. & Hilton J., 2003 — An exceptionally
preserved Lower Cretaceous ecosystem. Nature, 421: 807-
811.

Zinke J., 1998 — Small theropod teeth from the Upper Jurassic
coal mine of Guimarota (Portugal). Pa/dontologische
Zeitschrift, 72: 179-189.

Ziswiler V. & Farner D. S., 1972 — Digestion and the digestive
system. In: Avian Biology, Volume II. Farner D. S. & King
J. R. (eds.). Academic Press: 343-430.

Cristiano Dal Sasso
Museo Civico di Storia Naturale di Milano, Sezione di Paleontologia dei Vertebrati, Corso Venezia 55, 20121 Milano, Italia.
e-mail: cdalsasso@yahoo.com
Simone Maganuco
Museo Civico di Storia Naturale di Milano, Sezione di Paleontologia dei Vertebrati, Corso Venezia 55, 20121 Milano, Italia.
e-mail: simonemaganuco(@iol.it

Scipionyx samniticus (Theropoda: Compsognathidae) from the Lower Cretaceous of Italy
Osteology, ontogenetic assessment, phylogeny, soft tissue anatomy, taphonomy and palaeobiology
Memorie della Società Italiana di Scienze Naturali e del Museo Civico di Storia Naturale di Milano

Volume XXXVII - Fascicolo I

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 205

NOTE ADDED IN PROOF

Very recently, the anatomy of Juravenator was re-
described [Chiappe L. & Géhlich U., December 2010 —
Anatomy of Juravenator starki (Theropoda: Coelurosau-
ria) from the Late Jurassic of Germany. Neues Jahrbuch
fiir Geologie und Palciontologie, Abhandlungen, 258 (3):
257-296]. Given that this taxon is closely related to Sci-
pionyx and frequently mentioned throughout the present
monograph, and also that Scipionyx is frequently men-
tioned by Chiappe & Géhlich (2010), we thought it nec-
essary to add the following comparative remarks.

Nasal and lacrimal foramina — No nasal foramina are
reported in Juravenator, but the presence of at least two
of them is suggested by a UV photograph published by
Chiappe & Géhlich (2010: fig. 8B). Consistent with the
position they occupy in Scipionyx (Figs. 23-24, 27), these
foramina look to be craniocaudally aligned at the centre
of the nasals of Juravenator.

On the other hand, Chiappe & Géhlich (2010) con-
firm in writing a previous personal communication that
Juravenator, unlike Scipionyx, has a lacrimal foramen
(pneumatopore), not a simple depression.

Palatal bones — From the illustrations published by
Chiappe & Gòhlich (2010: fig. 8A-C) it can be seen that
some unnamed cranial bones, visible through the right in-
fratemporal, orbital and antorbital fenestrae of Juravena-
tor, are nearly in the same position as similarly shaped
bones that we identified in the skull of Scipionyx (Figs.
23-26), and likely represent homologous elements (e.g.,
right ectopterygoid).

Diastema — Chiappe & Géhlich (2010) contradict
themselves in writing that Scipionyx does (p. 263) and
does not (p. 268) possess a premaxillary-maxillary di-
astema. We confirm that such a diastema is present (Figs.
29, 44).

Tooth count and shape — Although Chiappe & Gòhlich
(2010) estimate a dentary tooth count for Scipionyx that is
wrong (12-14 contra 10; see Fig. 44), what is important
is that their estimate for Juravenator “did not exceed 11”.
In fact, a low number of dentary teeth, approaching that
present in Scipionyx, is consistent with the low number of
maxillary teeth in these two taxa (see also Low Number
Of Lateral Teeth in Ontogenetic Assessment).

With regard to tooth shape, we have well-document-
ed (see Heterodonty) that, contra Chiappe & Géòhlich
(2010), the teeth of Scipionyx are not “more homogenous
in shape”. Rather, like in Juravenator, the “maxillary and
posterior dentary teeth are more recurved than the pre-
maxillary teeth”.

Cervical pneumaticity — In the present monograph we
generically mention the presence of cervical pleurocoels
in Juravenator, having inferred this from the data matrix
of Gohlich & Chiappe (2006). In their redescription of
the specimen, Chiappe & Gòhlich (2010) specify that “the
third cervical in front of the first dorsal exhibits a small,
round foramen piercing the center of the centrum — this is
the only possible evidence of pneumaticity in the cervi-
cal series”. This position, relative to both the bone and

the vertebral series, is consistent with what is observed
in other basal coelurosaurs (see Cervical Vertebrae; Figs.
49-50, 53-54).

Neurocentral sutures — The lack of fusion between
sacral vertebrae and the presence of open neurocentral
sutures, visible on many caudal vertebrae of Juravena-
tor (Chiappe & Gohlich, 2010), is further evidence that in
compsognathids, as well as in other nonavian theropods,
the closure of the neurocentral sutures proceeds from the
cervical vertebrae in a caudal direction (see Incomplete
Ossification Of The Vertebral Column in Ontogenetic As-
sessment).

Furcula — In the new description of Juravenator, the
furcula is reported as “unfused” clavicles. The hatchling
Scipionyx, which is certainly more Juvenile than the speci-
men of Juravenator, unquestionably has two firmly sutured
elements (Figs. 87-89). Chiappe & Géhlich (2010) men-
tion this condition of Scipionyx and illustrate the clavicles
of Juravenator as clearly asymmetric (Chiappe & G6hlich,
2010: fig. 17). This, and their observation that one of the
two elements “exhibits a hook-like end that is possibly a
preservational artifact”, suggest that the furcula of Jurave-
nator, rather than being unfused, may have been not only
deformed but also broken by the diagenetic processes.

Carpal bones — According to Chiappe & Géhlich (2010),
Juravenator does not have any carpal bone, probably be-
cause not ossified yet — “another indication of the juvenile/
immature age of the specimen”. Seeing that ossified tarsal
elements are present in the hindlimb of the same specimen
(Chiappe & Géhlich, 2010), and that well-formed (radiale)
and even co-ossified (dc1+2) carpal bones are present in
the Scipionyx hatchling (Figs. 93-95), we wonder if the un-
labelled osseous elements seen under visible and UV light
in the well-articulated forelimb of Juravenator (Chiappe &
G6ohlich, 2010: fig. 20) may in fact be carpals.

Manual phalanges — In the present monograph (see
Manual Phalanges) we report that, according to Gòhlich
& Chiappe (2006, suppl. info., table I), digits I and III of
the manus of Juravenator are subequal in length, with the
latter digit slightly shorter than (i.e., 95% of the length
of) the former. In the new table of measurements (Chi-
appe & Gonhlich, 2010), the manual ungual of digit I is
now longer, so that digit III is now about 10% shorter than
digit I. This increases the difference between Juravenator
and Scipionyx in a character that may be autapomorphic
for Scipionyx (manual digit III markedly longer than digit
I is included in our emended diagnosis).

Reconstructions of Scipionyx — The new paper on
Juravenator contains three reconstructions of Scipionyx
that are based on Dal Sasso & Signore (1998a) and, con-
sequently, that are not up to date. Referring the reader to
the reconstructions included in this monograph, we would
like to point out that in the skull of Scipionyx redrawn by
Chiappe & Gòhlich (2010: figs. 9-10) there are portions
that in actual fact are not missing; that the mandibular fo-
ramen (i.e., the external mandibular fenestra) should be
omitted — correctly following the main text (Chiappe &

206 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Gohlich, 2010: p. 272); and that the small foramen in the
centre of the maxillary medial wall is actually a depres-
sion, not a perforation. Moreover, in the reconstruction of
Scipionyx°s forelimb (Chiappe & Géòhlich, 2010: fig. 19),
the scapular acromion should be not notched. In fact, in
the type specimen two patches of soft tissue overlap the
cranial margins of the left and right scapulae in a similar
manner, thus hiding scapular necks that gently merge into
the acromia (Figs. 87-88, 129).

Horny claws — The preservation of all the horny claws,
firmly attached to the manual and pedal ungual phalanges
of Juravenator, is remarkable and gives us the opportu-
nity to compare them with the homologous elements in
Scipionyx. The UV photographs published by Chiappe
& Génhlich (2010: fig. 20) show that the horny claws of
Juravenator have very similar shape, structure, curvature
and relative length to those of Scipionyx.

Internal soft tissue — Although the mode of preserva-
tion of Juravenator is different from that of Scipionyx,
we think that some remnants of internal soft tissue in the
two specimens may be homologous. Chiappe & Gòh-
lich (2010: fig. 24B) observe vertical strips in between
the chevrons of Juravenator and suggest that they reflect
segmentation of the tail. In our opinion, if these struc-
tures really lay on the medial sagittal plane, they are very
likely homologous to the similarly shaped laminae pre-
served between the chevrons of Scipionyx, that we refer
to the ligamentum interhaemale (Fig.164). On the other
hand, pending more detailed illustrations, we provision-
ally agree with Chiappe & Gòhlich (2010: fig. 24A) that
the structures paralleling the axis of the tail ventral to the
chevrons in Juravenator might be remnants of the septa of
the M. ilio-ischiocaudalis, although not so well-preserved
as in Scipionyx (Figs. 158-161).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 207

APPENDIX 1

Anatomical abbreviations used in text and figures
Abbreviazioni anatomiche riportate nel testo e nelle figure

der when preceding abbreviation, refer to left and right elements
davanti ad una abbreviazione, indicano rispettivamente elementi del lato sinistro e destro
() when including abbreviation, refer to elements overlain by others

quando racchiudono una abbreviazione, indicano elementi sottostanti

_ bold lines: visible limits of a given anatomical element

linee spesse: limiti visibili di un dato elemento anatomico

— thin lines: anatomical structures within a given element

linee sottili: strutture anatomiche all’interno di un dato elemento
--- hatched lines: estimated limits of an element overlain by others
linee tratteggiate: limiti stimati di un elemento sottostante

Italic latin names
Corsivo nomi in latino

Cranial skeleton / Cranio

af

ect

epcpt
eppt

ept

adductor fossa

fossa degli adduttori

angular

angolare

antorbital fenestra

finestra antorbitale

antorbital fossa

fossa antorbitale

ascending process of maxilla

processo ascendente del mascellare
articular

articolare

basisphenoid/parasphenoid
basisfenoide/parasfenoide

basal tuber

tubero basale

caudal process of squamosal

processo caudale dello squamoso
crista interfenestralis (interfenestral crest
cresta interfenestrale)

cultriform process

processo cultriforme

caudal palatine process of pterygoid
processo palatino caudale dello pterigoide
dentary

dentale

dentary teeth 1-10

denti 1-10 del dentale

tooth denticles

denticoli dei denti

dorsal margin of adductor fossa
margine dorsale della fossa degli adduttori
dorsal tympanic recess

recesso timpanico dorsale
ectopterygoid

ectopterigoide

apertura nasi ossea (external naris
narice esterna)

epipterygoid contact of pterygoid
contatto epipterigoideo dello pterigoide
ectopterygoid process of pterygoid
processo ectopterigoideo dello pterigoide
epipterygoid

epipterigoide

idp

frontal

frontale

fenestra ovalis (oval fenestra
finestra ovale)

fossa for ligamentum nuchae
fossa del legamento nucale
frontoparietal fontanelle
fontanella frontoparietale

hyoid

ioide

interdental plates

piastre interdentali

inner (ventral) wall of frontal
parete interna (ventral) del frontale
interfenestral bar

barra interfenestrale

inner (lingual) wall of maxilla
parete interna (linguale) del mascellare
inner (ventromedial) wall of nasal
parete interna (ventromediale) del nasale
internarial bar

barra internasale

interpterygoid vacuity

spazio interpterigoideo
infratemporal fenestra

finestra infratemporale

Jugal

giugale

jugal process of ectopterygoid
processo giugale dell’ectopterigoide
jugal process of maxilla

processo giugale del mascellare
jugal process of palatine

processo giugale del palatino
lacrimal

lacrimale

longitudinal bar of maxilla

barra longitudinale del mascellare
lateral condyle of quadrate
condilo laterale del quadrato
lateral process of pterygoid
processo laterale dello pterigoide
laterosphenoid

laterosfenoide

208

ofe

CRISTIANO DAL SASSO & SIMONE MAGANUCO

lacrimal vacuity

cavità lacrimale

maxilla

mascellare

maxillary teeth 1-7

denti mascellari 1-7

medial condyle of quadrate
condilo mediale del quadrato
maxillary fenestra

finestra mascellare

maxillary foramina

forami mascellari

Meckelian groove

solco di Meckel

medial glenoid of articular
glenoide mediale dell’articolare
maxillary medial wall

parete mediale del mascellare
mylohyoid notch

incisura miloioidea

maxillary process of palatine
processo mascellare del palatino
mandibular symphysis

sinfisi mandibolare

nasal

nasale

nasal foramina

forami nasali

orbital fenestra

finestra orbitale

orbitosphenoid

orbitosfenoide

parietal

parietale

palatine

palatino

premaxilla

premascellare

premaxillary teeth 1-5

denti premascellari 1-5
premaxillary foramina

forami premascellari
premaxillary process of vomer
processo premascellare del vomere
postorbital

postorbitale

paroccipital process

processo paroccipitale
postorbital process of parietal
processo postorbitale del parietale
postorbital process of squamosal
processo postorbitale dello squamoso
posteromedial process of pterygoid
processo posteromediale dello pterigoide
parietal process of squamosal
processo parietale dello squamoso
paraquadrate foramen

forame paraquadratico
prearticular

prearticolare

prefrontal

prefrontale

promaxillary fenestra

finestra promascellare

pro
prs

pt

prootic

prootico

promaxillary strut

puntello promascellare

pterygoid

pterigoide

pterygoid ala of quadrate

ala pterigoidea del quadrato
pterygoid process of ectopterygoid
processo pterigoideo dell’ectopterigoide
pterygoid process of palatine
processo pterigoideo del palatino
pterygopalatine process of vomer
processo pterigopalatino del vomere
quadrate

quadrato

quadrate ala of pterygoid

ala quadratica dello pterigoide
quadrate cotyle of squamosal

cotilo quadratico dello squamoso
quadrate head

testa del quadrato

quadratojugal

quadratogiugale

quadratojugal contact of quadrate
contatto quadratogiugale del quadrato
quadratojugal process of squamosal
processo quadratogiugale dello squamoso
retroarticular process

processo retroarticolare

rostral ramus of maxilla

ramo rostrale del mascellare
surangular

soprangolare

subtemporal fenestra

finestra sottotemporale

scleral plates

placche della sclera

subnarial foramen

forame sottonasale

supraoccipital

sopraoccipitale

suborbital fenestra

finestra sottorbitale

splenial

spleniale

subsidiary palatal fenestra

finestra palatale sussidiaria
squamosal

squamoso

sinusoidal ridge of supratemporal fossa

cresta sinusoidale della fossa sopratemporale

sagittal suture

sutura sagittale

supratemporal fenestra

finestra sopratemporale

subnarial process of nasal

processo sottonasale del nasale
subnarial process of premaxilla
processo sottonasale del premascellare
supranarial process of premaxilla
processo sopranasale del premascellare
transverse nuchal crest of parietal
cresta nucale trasversale del parietale

ved
vfect
vmaf
vppal
vppt

wmfe

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 209

transverse ridge of supraoccipital
cresta trasversale del sopraoccipitale
vomer

vomere

entrance of vena capitis dorsalis
ingresso della vena dorsale del capo
ventral fossa of ectopterygoid

fossa ventrale dell’ectopterigoide
ventral margin of adductor fossa
margine ventrale della fossa degli adduttori
vomeral process of palatine
processo vomerale del palatino
vomeral process of pterygoid
processo vomerale dello pterigoide
osseous wall of maxillary fenestra
parete ossea della finestra mascellare

Axial skeleton / Scheletro assiale

asil
assr
ati

atn

cap

articular surface for ilium
superficie articolare per l’ileo
articular surface for sacral rib
superficie articolare per la costola sacrale
atlantal intercentrum
intercentro dell’atlante
atlantal neurapophysis
neurapofisi dell’atlante

axial intercentrum
intercentro dell’epistrofeo
axial neural spine

spina neurale dell’epistrofeo
cervical vertebra

vertebra cervicale

caudal vertebra

vertebra caudale

caudal articular facet

faccetta articolare caudale
caudal centrum

centro caudale

caudal neural arch

arco neural caudale
capitulum

capitello

cervical centrum

centro cervicale

costal groove

solco costale

craniolateral process
processo craniolaterale
concavity of the neural spine
concavità della spina neurale
cervical rib

costola cervicale

cranial articular facet
faccetta articolare craniale
cranial process

processo craniale

cranial chevron-shaped gastralium
gastralium craniale a forma di V
capitulotubercular web
membrana capitello-tubercolare
dorsal vertebra

vertebra dorsale

dorsal centrum

centro dorsale

dna

tpolf

tpolp

dorsal neural arch

arco neurale dorsale
diapophysis

diapofisi

dorsal rib

costola dorsale

epipophysis

epipofisi

gastralia

haemal arch (chevron)

arco emale

interspinal ligament attachment
inserzione dei legamenti interspinali
infrapostzygapophyseal fossa
fossa infrapostzigapofisaria
lateral gastralium

gastralium laterale
mediodorsal facet

faccetta mediodorsale

medial gastralium

gastralium mediale
medioventral facet

faccetta medioventrale

neural canal

canale neurale

neurocentral suture

sutura neurocentrale
neurocentral articular surface
superficie articolare neurocentrale
neural spine

spina neurale

posterior centrodiapophyseal lamina
lamina centrodiapofisaria posteriore
pneumatopore

pneumatoporo
postzygodiapophyseal lamina
lamina postzigodiapofisaria
postzygapophysis
postzigapofisi
paradiapophyseal lamina
lamina paradiapofisaria
prezygodiapophyseal lamina
lamina prezigodiapofisaria
prezygoepipophyseal lamina
lamina prezigoepipofisaria
prezygoparapophyseal lamina
lamina prezigoparapofisaria
prezygapophysis

prezigapofisi

sacral vertebra

vertebra sacrale

sacral centrum

centro sacrale

sacral neural arch

arco neurale sacrale

sacral rib

costola sacrale

transverse process

processo trasverso
intrapostzygapophyseal fossa
fossa infrapostzigapofisaria
intrapostzygapophyseal pneumatopore
pneumatoporo infrapostzigapofisario

210

tu

wemvf wing-like expansion of the medioventral facet
espansione ad ala della faccetta medioventrale

Appendicular skeleton / Scheletro appendicolare

I-II

1-4

ac
actil

actis

bf
ca

cabil

ccil

Gli

tuberculum
tubercolo

first to third digit

dito dal primo al terzo

first to fourth phalanx

falangi dalla prima alla quarta
acromion

acetabular portion of ilium
porzione acetabolare dell’ileo
acetabular portion of ischium

porzione acetabolare dell’ischio

brevis fossa

carpals

carpali

caudal blade of ilium
lama caudale dell’ileo
cranial concavity of ilium
concavità craniale dell’ileo
fossa of collateral ligament
fossa del legamento collaterale
cnemial crest

cresta cnemiale

coracoid

coracoide

coracoid foramen

forame coracoideo
coracoid tubercle
tubercolo del coracoide
cranial blade of ilium
lama craniale dell’ileo
cranioventral process
processo cranioventrale
distal carpals 1+2

carpali distali 1+2
deltopectoral crest

cresta deltopettorale
epicleideum

extensor pit

fossa dell’estensore
femur

femore

fibula

fibular condyle

condilo fibulare

flexor tubercle

tubercolo del flessore
furcula

symphysis of furcula
sinfisi della furcula
glenoid fossa

fossa glenoidea

greater trochanter
trocantere maggiore
humeral head

testa dell’omero

hooked process of ilium
processo uncinato dell’ileo
humerus

omero

hypocleideum

il
iis
ilpu
inc
is
isf
ismf
isp

iti
lec

It
me
mec
mfi
obn
obp
olp
pfo
phar
pu
pua
puf
pufo
PUp
ra

rae
re

sac
sca
scono
SÌ

ti

uc

ul

CRISTIANO DAL SASSO & SIMONE MAGANUCO

ilium

ileo

iliac process of ischium
processo iliaco dell’ischio
iliac process of pubis
processo iliaco del pube
internal condyle
condilo interno

ischium

ischio

ischial foot

piede ischiatico

ischial medial facet
faccetta mediale dell’ischio
ischial peduncle
peduncolo ischiatico
incisura tibialis

lateral epicondyle
epicondilo laterale
lesser trochanter
trocantere minore
metacarpal

metacarpale

medial epicondyle
epicondilo mediale
medial fossa of fibula
fossa mediale della fibula
obturator notch

incisura otturatoria
obturator process
processo otturatore
olecranon process
processo olecranico
popliteal fossa

fossa poplitea
phalangeal articulation
articolazione della falange
pubis

pube

pubic apron

grembiule pubico

pubic foot

piede pubico

pubic foramen

forame pubico

pubic peduncle
peduncolo pubico
radius

radio

radiale

radial condyle

condilo radiale
supracetabular crest
cresta sopracetabolare
scapula

scapola
scapulocoracoidal notch
incisura scapolocoracoidea
spatium interosseum
tibia

ulnar condyle

condilo ulnare

ulna

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Soft anatomy / Tessuti molli

Ab
aduo
alt
ana
apa
atrar
ba

bv

iem
ilem
11Cms
ile

ill

int

ipam

A band

banda A

ascending loop of the duodenum
ansa ascendente del duodeno
amorphous indeterminate tissue
tessuto amorfo indeterminato
anastomoses

anastomosi

apatite crystallites

cristalliti di apatite

dorsal apexes of one tracheal ring
apici dorsali di un anello tracheale
bacteria

batteri

blood vessel

vaso sanguigno

branch of blood vessel
ramificazione del vaso sanguigno
blood vessel wall

parete del vaso sanguigno
canaliculi

canalicoli

M. caudofemoralis longus (caudofemoral muscle
muscolo caudofemorale)
collagen bundles

fasci di collagene

connective tissue

tessuto connettivo

descending loop of the duodenum
ansa discendente del duodeno
duodenum

duodeno

endomysium and sarcolemma
endomisio e sarcolemma
oesophagus

esofago

fibrous articular cartilage
cartilagine articolare fibrosa
faecal pellet

massa fecale

fish scales

squame di pesce

hyaline articular cartilage
cartilagine articolare ialina

horny claw

artiglio corneo

I band

banda I

M. ischiocaudalis (ischiocaudal muscle
muscolo ischiocaudale)

M. ilio-ischiocaudalis (ilio-ischiocaudal muscle
muscolo ileo-ischiocaudale)
ilio-ischiocaudal muscle septa
setti del muscolo ileo-ischiocaudale
ileum

ileo

interlamellar line

linea interlamellare

intestine

intestino

hypaxial musculature
muscolatura ipoassiale

Je]

lam
lih

liv

lu
mead
mes
myo
ncare
ol
pifm
plci
psila
SG
sar
sto
sun
tra
trar
ts

un

211

Jjejunum

digiuno

lamella

ligamentum interhaemale (interhaemal ligament
legamento interemale)

liver and other blood-rich organs

fegato e altri organi ricchi di sangue
lumen

lume

muscularis externa and/or adventitia
tonaca muscolare esterna e/o avventizia
mesenteric connection

connessione mesenterica

myofiber

miofibra

neurocentral articular cartilage
cartilagine articolare neurocentrale
osteocyte lacuna

lacuna osteocitaria

puboischiofemoral muscle

muscolo puboischiofemorale

plicae circulares (circular folds

pieghe circolari)

posterior sacroiliac ligament attachment
inserzione del legamento sacroiliaco posteriore
rectum

retto

sarcomere

sarcomero

stomach contents

contenuto stomacale

subunguis

trachea

tracheal ring

anello tracheale

transverse section

sezione trasversa

unguis

Chemical elements / Elementi chimici

AI
Au

C

Si

Aluminium (aluminium
alluminio)

Aurum (gold

Oro)

Carbonium (carbon
carbonio)

Calcium (calcium
calcio)

Ferrum (iron

ferro)

Natrium (sodium
sodio)

Oxygenum (oxygen
ossigeno)

Phosphorus (phosphorus
fosforo)

Silicium (silicon
silicio)

2A | 9) CRISTIANO DAL SASSO & SIMONE MAGANUCO

APPENDIX 2
Modifications to the character list of Senter (2007)

The characters of Senter (2007; and references therein)
are not repeated here, except where changes to the char-
acter definitions have been made or character states have
been added/deleted. Details of the former description are
provided in square brackets, with accompanying explana-
tions for changes where needed.

ch27. Pronounced accessory antorbital fenestra absent (0)
or present (1) [Formerly: Pronounced, round accessory
antorbital fenestra absent (0) or present (1). Reworded
according to the observation that in some taxa the
accessory antorbital fenestra is not round].

ch28. Accessory antorbital fenestra situated at rostral
border of antorbital fossa (0) or situated caudal to rostral
border of fossa (1) [Formerly: Accessory antorbital fossa
situated at rostral border of antorbital fossa (0) or situated
caudal to rostral border of fossa (1)].

ch32. Postorbital process of the jugal: well-developed,
taller than half orbit (0) or reduced/absent (1) [Former-
ly: Jugal and postorbital contribute equally to postor-
bital bar (0) or ascending process of jugal reduced and
descending process of postorbital ventrally elongate
(1). In the data matrix of Senter (2007) there were no
taxa coded (1), so that the previous character was un-
informative].

ch45. Frontal edge smooth in region of lacrimal su-
ture (or prefrontal suture, where the contact is with
this bone) (0) or edge notched (1) [Formerly: Frontal
edge smooth in region of lacrimal suture (0) or edge
notched (1). Reworded according to the observation
that in Scipionyx the frontal edge contacts the prefron-
tal. Scipionyx is coded (0), the edge forming a faint,
unnotched concavity].

ch48. Descending process of squamosal nearly parallels
quadrate shaft (0) or nearly perpendicular to quadrate
shaft (1) [Formerly: Descending process of squamosal
parallels quadrate shaft (0) or nearly perpendicular to
quadrate shaft (1)].

ch63. Palatine and ectopterygoid separated by pterygoid
or making point contact, at most (0) or contact (1) [For-
merly: Palatine and ectopterygoid separated by ptery-
goid (0) or contact (1). In Scipionyx, palatine and ectop-
terygoid make point contact, but they can be considered
as separated].

ch73. External mandibular fenestra oval (0), subdivided
by a spinous rostral process of the surangular (1), or absent
(2) [Formerly: External mandibular fenestra oval (0) or
subdivided by a spinous rostral process of the surangular
(1). We added the character state (2), several taxa lacking
a mandibular fenestra].

ch74. Internal mandibular fenestra absent or small and
slit-like (0) or large and rounded (1) [Formerly: Internal
mandibular fenestra small and slit-like (0) or large and
rounded (1). Reworded according to the observation

that in some taxa (e.g., Compsognathus) the internal
mandibular fenestra can be considered as absent; this
absence potentially represents a new derived character
state, however, it can not be ascertained pending new,
more complete specimens, and new studies on the known
ones].

ch77. Coronoid ossification present (0) or absent (1)
[Formerly: Coronoid ossification large (0) or only a thin
splint (1) or absent (2). The coronoid is the complex
supradentary-coronoid (e.g., Holtz et a/., 2004)].

ch101. Cervical and cranial trunk vertebrae amphiplatyan
to platycoelous (0) or strongly opistocoelous(1)[Formerly:
Cervical and anterior trunk vertebrae amphiplatyan (0) or
opisthocoelous (1)].

ch106. Dorsal vertebrae not pneumatic (0) or pneumatic
(1) [Formerly: Middle and posterior dorsal vertebrae not
pneumatic (0) or pneumatic (1)].

ch109 Scars for interspinal ligaments terminate at apex
of neural spine in dorsal vertebrae (0) or terminate be-
low apex of neural spine (1), or interspinal ligament
attachments in form of beak-like processes below apex
of neural spine (2) [Formerly: Scars for interspinal
ligaments terminate at apex of neural spine in dorsal
vertebrae (0) or terminate below apex of neural spine
(1). Character state (2) has been added to reflect the
condition visible in some basal coelurosaurus, Scipio-
nyx included].

ch154. Ventral edge of cranial ala of ilium straight or
gently curved (0) or ventral edge hooked cranially (1)
[Formerly: Ventral edge of anterior ala of ilium straight
or gently curved (0) or ventral edge hooked anteriorly (1)
or very strongly hooked (2). Character state (2) has been
deleted because there are no taxa showing that character
state in the present data set].

ch168. Obturator process of ischium contacts pubis in
the distal half (1) or not (0) [Formerly: Obturator process
does not contact pubis (0) or contacts pubis (1)].

ch169. Craniocaudal length of the caudal process of
the pubic foot: process absent (0); process present but
shorter than 1/5 of the proximodistal length of the pu-
bis (1); process present and long more than 1/5 but
less than 1/3 of the proximodistal length of the pubis
(2); process present and longer than 1/3 of the proxi-
modistal length of the pubis (3) (Modified from Holtz,
2000; following Cau, pers. comm., 2009). [Formerly:
Length of pubic boot <30% of length of pubis (0) or
>40% (1). There are some taxa (e.g., Sinocalliopteryx)
showing intermediate values that render impossible
the application of the character as proposed by Senter
(2007), and render difficult, at the same time, to define
new values for separating the character states; for this
reason we have preferred to focus our attention on the
caudal process of the pubic foot, and to change the
whole character].

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 213

ch205. Foot symmetrical (0), asymmetrical with slender
mt II and very robust mt IV (1), or asymmetrical with
mt IV reduced in mediolateral width (2) [Formerly: Foot
symmetrical (0), or asymmetrical with slender mt Il
and very robust mt IV (1). Character state (2) has been
introduced to describe the condition in some taxa, such as
Nqwebasaurus and Aniksosaurus].

ch207. Shaft diameter of phalanx I-1 less (0) or equal/
greater (1) than shaft diameter of radius [Reworded
because there are some taxa in which phalanx I-1 is equal
to the shaft diameter of the radius, resulting more similar
to the condition in which phalanx I-1 is greater than the
shaft diameter of the radius, than to the opposite condition.
Formerly: Shaft diameter of phalanx I-1 less (0) or greater
(1) than shaft diameter of radius].

ch231. Dentary teeth do not (0) or do (1) increase in
size rostrally [Formerly: Dentary teeth do not (0) or do
increase in size rostrally, becoming more conical in shape
(1). Senter (2007) commented: “Character reworded
to refer only to tooth size”. He maintained, however, a
reference to the tooth shape; we completely deleted any
reference to the tooth size to avoid confusion - €.g.,
Compsognathus should be coded (0) for tooth size and (1)
for tooth shape].

ch235. Cranial concavity of the preacetabular blade of
the ilium in lateral view: absent (0); present and slightly
developed (1); present and craniocaudally expanded
(2) [Formerly: Snout does not (0) or does taper to an
anterior point (1). This character from Senter (2007) has
been excluded because equivocal: for example, it is not
clear in which view the snout is intended to taper, and
why Archaeopteryx is coded (1) and Compsognathus
(0) in the Senter (2007) data matrix. The new character
focusses on the cranial concavity of the preacetabular
blade of the ilium seen in several basal coelurosaurs; the
same concavity is described as “craniodorsal” in Rauhut
(2003), and “anterodorsal” in Xu et a/. (2004); although
it is virtually cranioventral in position (Carpenter et al.,
2005b: fig. 3.10), the concavity shown by Ornitholestes
was coded (1), supposing it as homologous].

ch239. Rostral portion of the maxillary antorbital fossa:
small, from 10% to less than 40% of the rostrocaudal
length of the antorbital cavity (0), large, greater than
40% of the rostrocaudal length of the antorbital cavity (1)
[Formerly: Maxillary antorbital fossa: small, from 10% to
less than 40% of the rostrocaudal length of the antorbital
cavity (0), large, greater than 40% of the rostrocaudal
length of the antorbital cavity (1)].

ch240. Maxillary fenestra large and round (0), a large,
craniocaudally elongate oblong(1), asmall, craniocaudally
elongate slit, not dorsally displaced (2), a small, dorsally
displaced opening (3), or a small and round, not dorsally
displaced opening (4) [Formerly: ch240. Maxillary
fenestra large and round (0), a large, craniocaudally
elongate oblong (1), a small, craniocaudally elongate slit,
not dorsally displaced (2), or a small, dorsally displaced
opening (3). Character state (4) has been added to reflect
the condition in Juravenator and Huaxiagnathus].

ch270. Acromion process does not match any of the
following descriptions (0), rectangular with its dorsal edge
forming a 90° angle with the dorsal edge of the scapular
blade (1) or a quarter-circle in shape (2) or triangular,
with apex pointing away from and subparallel to scapular
blade (3) [Formerly: Acromion process does not match
any of the following descriptions: (0) rectangular with its
dorsal edge forming a 90° angle with the dorsal edge of
the scapular blade (1) or a quarter-circle in shape (2) or
triangular, with apex pointing away from and subparallel
to scapular blade (3)].

214 CRISTIANO DAL SASSO & SIMONE MAGANUCO

APPENDIX 3
Major changes in character codings with respect to the data matrix of Senter (2007)

ch40. Thishas been coded(2) for mostofthe ornithomimids
in which the character can be scored (see Data matrix in
Appendix 6), with the only exception of Pel/ecanimimus
polyodon. This is contra Senter (2007), in which the
character state for most ornithomimids is (0).

ch47. The codings have been checked for all the taxa,
with the following additions: Harpymimus okladnikovi
(0); Caudipteryx (1); Oviraptor philoceratops (1); and
Epidendrosaurus ningchengensis (1).

ch51. The codings have been checked for all the taxa,
with the following additions: Eotyrannus lengi (0);
Gorgosaurus libratus (0), Garudimimus brevipes (0);
Erlikosaurus andrewsi (0), Protarchaeopteryx robusta
(0); Epidendrosaurus ningchengensis (0); Confuciusornis
sanctus (1); and Yanornis martini (1).

ch255. This character has been coded as inapplicable (-)
in taxa lacking an external mandibular fenestra.

ch258. This character has been coded as inapplicable (-)
in taxa lacking mesial denticles, including the troodontids
Byronosaurus jaffei and Troodon formosus, which
formerly were coded (1).

Huxiagnathus orientalis

Additions: ch37(0); ch73(2); ch85(0/2); ch96(0);
ch107(0); ch109(2); ch116(0); ch117(0); ch121(0/1);
ch187(0); ch192(0); ch204(0); ch268(0); ch282(1);
ch346(1); ch357(0).

Changes: ch42(2); ch115(0); ch126(0); ch142(0/1);
ch175(1); 202(?); ch236(?); ch240(4); ch251(?); ch255(-);
ch258(-); ch263(1); ch283(1); ch284(1); ch286(1);
ch299(1); ch360(-).

Compsognathus longipes

Additions: ch20(0); ch22(0); ch28(1); ch29(1); ch31(0);
ch32(0); ch33(1); ch34(1); ch41(1); ch42(2); ch45(0);
48(0); 49(0); 50(0); 51(0); 52(0); 73(2); 74(0); 76(0);

77(0); 78(0); 80(0); ch101(1);
ch110(0); ch127(0); ch134(0);
ch146(1); ch148(0); ch149(0);
ch153(0); ch154(0); ch156(0);
ch193(0); ch194(1); ch210(0);
ch225(0); ch226(0); ch227(0);
ch257(0/1); ch260(0); ch268(0);
ch283(0); ch284(1); ch285(0);
ch288(1); ch289(0); ch290(0);
ch293(0); ch298(1); ch328(0);
ch345(0); ch354(0); ch360(-).

ch107(0);
ch138(0);
ch150(0);
ch174(0);
ch214(0);
ch241(0);
ch280(0);
ch286(1);
ch291(0);
ch337(0);

ch109(2);
ch141(0);
ch152(0);
ch192(0);
ch218(0);
ch256(0);
ch282(1);
ch287(0);
ch292(1);
ch344(0);

Changes: ch71(0); ch96(0); ch115(0); ch126(0); ch135(?);

ch181(0); ch182(0); ch184(0);
ch258(-); ch263(1); ch275(1).

ch233(1):

ch255(-);

Sinosauropteryx prima

Additions: ch33(0); ch69(0); ch76(0); ch109(2); ch160(0);
ch204(0); ch226(1); ch227(0).

Changes: ch73(2); ch96(0); ch126(0); ch175(1); ch202(?);
ch255(-); ch258(-); ch263(1).

Dilong paradoxus
Additions: ch51(1); ch256(1).
Changes: ch72(1).

Tanycolagreus topwilsoni
Additions: ch146(1); ch211(0).
Changes: ch148(0); ch271(0); ch310(1).

Ornitholestes hermanni
Additions: ch84(1).
Changes: ch109(2);
ch235(1); ch314(0).

ch117(1); ch1620); chivs30

Coelurus fragilis
Additions: ch117(1).
Changes: ch100(0/1); ch169(?); ch263(0/1); ch310(1).

Bambiraptor feinbergi
Changes: 135(?); 241(2).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 215

APPENDIX 4
Comments on selected characters of Senter (2007)

ch43. Anterior emargination of supratemporal fossa
on frontal straight or slightly curved (0) or strongly
sinusoidal and reaching onto postorbital process (1). In
Scipionyx the fossa is confined to the lateral part of the
skull roof, showing a limited mediolateral extension;
the caudal margin of the frontal, however, forms a
sinusoidal emargination reaching the postorbital process
in correspondence with the fossa and well-extended
medially, farther than the medial extension of the fossa.
Taking into account the whole sinusoidal emargination,
and not only its portion in correspondence with the fossa,
Scipionyx is coded (1).

ch89. Dentary teeth evenly spaced (0) or rostral
dentary teeth smaller, more numerous and more closely
appressed than those in middle of tooth row (1). The
dentary teeth of Scipionyx are evenly spaced from each
other but slightly more closely appressed than those
in the middle of the tooth row. Nevertheless, they are
neither more numerous nor smaller, being rather larger
than those in the middle of the tooth row. Therefore,
Scipionyx can be coded (0).

ch108. Neural spines of dorsal vertebrae not expanded
distally (0) or expanded to form a “spine table” (1). In
Scipionyx the apex of the neural spine in dorsal vertebrae
2,3 and 6 is feebly expanded, but in our opinion it cannot
be considered a true spine table and Scipionyx has been
coded (0). However, this character should be more clearly
quantified.

ch142. Olecranon process weakly developed (0)
or distinct and large but not hypertrophied (1) or
hypertrophied (2). Character state (2) is clearly referred
to the hypertrophied olecranon process of alvarezsaurids.
On the other hand, distinction between character states
(0) and (1) is not so clear in certain cases, some taxa
showing “intermediate” conditions, and the state (1), as
it is, includes a large variability of development of the
olecranon process. For example, Scipionyx, as well as
Sinosauropteryx and Compsognathus, can be coded (1).
However, the olecranon process of Sinosauropteryx is
comparatively more developed than that of Scipionyx
and Compsognathus, being sometimes described as
“hypertrophied”, although it is not as hypertrophied as in
alvarezsaurids. This degree of variability would require
more states for this character, more clearly separated
and, possibly, well-quantified.

ch148. Distal carpals 1+2 well-developed, covering all
of proximal ends of metacarpals I and II (0) or small,
cover about half of base of metacarpals I and II (1) or
cover bases of all metacarpals (2). Scipionyx can be
coded (1), although it shows a particular condition (see
Forelimb) in which most of the proximal end of Mel is
covered. Huaxiagnathus and Sinocalliopteryx have been
coded (1) as well, their distal carpals 1+2 contacting
both MclI and MclI, although covering less than half of
the two bones.

ch206. Neural spines on caudal dorsal vertebrae in lateral
view rectangular or square (0) or craniocaudally expanded
distally, fan-shaped (1). Scipionyx has been coded (1), but
see description (Dorsal Vertebrae) about the use of the
term “fan-shaped”.

216 CRISTIANO DAL SASSO & SIMONE MAGANUCO

APPENDIX 5
Comments on the coding of character states possibly affected by juvenile ontogenetic stage

ch2. Orbit round in lateral or dorsolateral view (0) or
dorsoventrally elongate (1). Among basal coelurosaurs, a
dorsoventrally elongate orbit is present only in the adults
of the large-sized tyrannosaurids 7yrannosaurus and
Gorgosaurus. All the other taxa maintain a round orbit in
all growth stages; therefore, Scipionyx has been coded (0).

ch38. Supraorbital crests on lacrimal in adult individuals
absent (0) or dorsal crest above orbit (1) or lateral
expansion rostral and dorsal to orbit (2). Dorsal crests and
rugosities above orbit and lateral to it are often present
in allosauroids, but are uncommon in coelurosaurs (e.g.,
Holtz et al., 2004). A lateral expansion rostral and dorsal to
the orbit is present in the ornithomimosaurs Garudimimus
and Pelecanimimus, and in the troodontids 7roodon,
Byronosaurus and Zanabazar. All these taxa are known
from adult individuals. In Scipionyx, the lacrimal closely
resembles that of other basal coelurosaurs, both juveniles
and adults. For this reason, we have coded Scipionyx (0),
pending adult material.

ch46. Dorsal surface of parietals flat, lateral ridge
borders supratemporal fenestra (0) or parietals dorsally
convex with very low sagittal crest along midline (1) or
dorsally convex with well-developed sagittal crest (2).
Parietal crests are very low or absent in various small-
sized theropods not strictly related to each other, such as
coelophysoids, Compsognathus, ornithomimosaurs and
oviraptorosaurs. In ornithomimosaurs (Makovicky et al.,
2004), the sagittal crest is absent in both juveniles and
adults. A sharp but rather low sagittal ridge is present in
some adult oviraptorosaurs (Osmélska et al., 2004) but is
totally absent in early ontogenetic stages of Citipati and
cf. Rinchenia (Auditore, pers. comm., 2009), in which,
however, the parietals are already convex dorsally.
A well-developed sagittal crest is present in both
tyrannosauroids and some troodontids. In tyrannosaurids
the sagittal crest is already well-developed in juveniles
(Carr & Williamson, 2004). Tykoski (2005) did not
include this character in the listofthe maturity-dependent
characters deleted from the analysis. In all the above
mentioned taxa, the parietal does not vary noticeably
in curvature during ontogeny. Concerning Scipionyx,
the sagittal crest is absent, and the dorsal surface of the
parietal is almost flat (see Parietal), with lateral ridges
bordering extensively the supratemporal fenestra. Based
on all these data, we provisionally coded Scipionyx (0),
pending adult material.

ch47. Parietals separate (0) or fused (1). In Scipionyx the
parietals are separated. However, this character is often
indicated as a maturity-dependent character (e.g., Tykoski,
2005). For this reason it has been coded (?).

ch6éS. Suborbital fenestra similar in length to orbit (0) or
about half or less than half orbital length (1) or absent
(2). The orbit in Scipionyx is very large, more than four
times the length of the suborbital fenestra. However, the
size of the orbit is clearly maturity-dependent, as is often
the case in extant vertebrates. For this reason, Scipionyx
is coded (0/1).

ch85. Dentary and maxillary teeth large, less than 25
in dentary (0) or large number of small teeth (25 or
more in dentary) (1) or small number of dentary teeth
(<11) (2) or dentary without teeth (3). Scipionyx has
10 dentary teeth. However, as the formation of new
tooth positions would be expected during growth (see
Ontogenetic assessment), Scipionyx has been coded

(2).

ch135. Scapula and coracoid separate (0) or fused in-
to scapulocoracoid (1). According to Tykoski (2005),
scapula and coracoid fuse in adults. For this reason,
Scipionyx, Juravenator, Compsognathus and Bambi-
raptor have been scored (?).

ch139. Scapula longer than humerus (0) or humerus
longer than scapula (1). As the girdles are compara-
tively small in size in immature theropods (see Ontoge-
netic Assessment), Scipionyx has been coded (?).

ch169. Craniocaudal length of the caudal process of
the pubic foot: process absent (0); process present but
shorter than 1/5 of the proximodistal length of the pu-
bis (1); process present and long more than 1/5 but less
than 1/3 of the proximodistal length of the pubis (2);
process present and longer than 1/3 of the proximodis-
tal length of the pubis (3). According to Tykoski (2005)
this character is maturity-dependent. In Scipionyx a
caudal process of the pubic foot is present, and it is
craniocaudally shorter than 1/5 of the proximodistal
length of the pubis. However, we cannot exclude that
during ontogeny it would have become longer. For this
reason Scipionyx is coded (1/3).

ch232. Length of skull more than 90% of femoral
length (0) or less than 80% (1). Most basal coeluro-
saurs have a skull length that is more than 90% of the
femoral length. In Scipionyx the skull length is 130%
of the femoral length. The skull is longer than the fe-
mur also in Juravenator and in the large, more mature
compsognathids Sinocallipteryx and Huaxiagnathus,
and it approximates femoral length in Compsognathus.
In both juveniles and adults of 7yrannosaurus (Brochu,
2003; Mortimer, 2004-2010) the skull is almost as long
as the femur. In juveniles and adults of Sinornithom-
imus (Kobayashi & Lii, 2003) the skull is 61% and 69%
of the femoral length, respectively. Taking into account
those examples, it is very unlikely to suppose that in
Scipionyx the skull, starting from 130%, would have
become shorter than 80% of the femoral length during
growth. For this reason Scipionyx is coded (0).

ch236. Area of antorbital fenestra greater than that
of orbit (0) or less than that of orbit (1). In non-
tyrannosauroid basal coelurosaurs the area of the
antorbital fenestra measures less than the orbital area.
In Scipionyx the snout should have become more than
three times longer than it is, in order to have an area of
antorbital fenestra comparable to that of the orbit. For
this reason Scipionyx is coded (1).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY Dez

ch239. Rostral portion of the maxillary antorbital fossa:
small, from 10% to less than 40% of the rostrocaudal
length of the antorbital cavity (0), large, greater than 40%
of the rostrocaudal length of the antorbital cavity (1). The
rostral portion of the maxillary antorbital fossa is very
small in Scipionyx, due to the shortness of the snout. As
this character is clearly maturity-dependent, the Italian
compsognathid has been coded (?).

ch240. Maxillary fenestra large and round (0), a large,
craniocaudally elongate oblong(1), a small, craniocaudally
elongate slit, not dorsally displaced (2), a small, dorsally
displaced opening (3), or a small and round, not dorsally
displaced opening (4). The snout in Scipionyx is very
short craniocaudally and steeply inclined, leaving little
space for the maxillary fenestra. We cannot exclude
that the small, dorsally displaced opening of Scipionyx
would have changed in shape and position during growth,
therefore this taxon has been coded (?).

ch241. Nasal fusion: absent, nasals separate (0) or present,
nasals fused together (1). According to Tykoski (2005) this
character is maturity-dependent, and has been coded (?) in
taxa not represented by adult individuals (e.g., Scipionyx,
Juravenator, Compsognathus and Bambiraptor). The only
exception is Di/ong, in which the nasal fusion occurred
prior to maturity (Xu et al., 2004), so that it has been
coded (1).

ch244. Nasals at least as long as frontals (0) or shorter
than frontals (1). In Scipionyx the nasal is shorter than
the frontal (1). However, the condition in all likelihood
reflects the early ontogenetic stage of the individual (see
Ontogenetic Assessment).

DAR CRISTIANO DAL SASSO & SIMONE MAGANUCO

APPENDIX 6
Data matrix

Taxa / Characters 1 2 3 4 5 6 7 8 GIL LITNI2I8I3 LAM
Allosaurus fragilis i; ai gi 0 0 0 fe 0 0 0 0 0 0 0 0
Sinraptor ? di gl 0 0 0 È ? 0 0 ? 0 0 0 0
Dilong paradoxus ? p 1 0 0 Vi ? 2 E; f 0 0 0 ? È
Eotyrannus lengi f vi È. Di ? d t; 5 È; È 6 id 5; ti vi
Tyrannosaurus rex È al 0 0 0 0 P (0) 0 ? 0 (0) 0 0 0
Gorgosaurus libratus ? al 0 0 0 0 2 0 0 E 0 0 0 ? (0)
Tanycolagreus topwilsoni E a 0 0 ti ? ? 2 Di 2 ? ? È ot; cd;
Coelurus fragilis H fi; G; Gi ? f; l ? t 5 ? ? Li R d
Ornitholestes hermanni e 0 al al 0 È o ? (0) È (0) P 0 0 i
Guanlong wucaii È 0 1 0) 0 Mi ? PIO LE tl: È È ? e (;
Aniksosaurus darwini ® E 5) ? 2 e 2 ? f) ? È ? ? È 2
Nedcolbertia justinhoffmani è ? ? ? ? ? ip, 2 P; ? È R ? e È
Nqwebasaurus thwazi ? 0 ? E È È ? È i; ? ? CON ? ?
Santanaraptor placidus ? n ? È ? fo di 2 E È 6 6 2 fs ?
Orkoraptor burkei È L. al 1 ? 2 È Ci F; E; di G P ? ?
Juravenator starki ? 0 Il 1 0 ? D: 2, P ? p ? Gi 2 D
Mirischia asymmetrica i; i; ? ? ? ? ? E; ? 2 ? fi di ? PR
Compsognathus longipes ? 0 È nl (0) D a ? E ? e, ? Ls Di 2
Sinocalliopteryx gigas %; 0 1 fi 5 w ? È 2 = G; P © ? ?
Huaxiagnathus orientalis ? 0 Hi ? 0 E Gi ? 1 ? Li 9 È fi È
Sinosauropteryx prima È 0 ali ? 0 E ? ? € wi ? fi PI 2 2
Scipionyx samniticus D; 0 al dl O) ? 0 1 2 ? 0 È : ? e)
Deinocheirus mirificus È G È #; ? ? È ? 6; R P ? ? te ;
Harpymimus okladnikovi ? 0) il I ? E È; ? È E; È ? ? È ?
Pelecanimimus polyodon sE 0 Il e 6 Fi ih ? P ? 1 ta È ? p
Shenzhousaurus orientalis P. ? ? È ? ? ? fi E ? L È È E; ?
Archaeornithomimus asiaticus E te ? ? f È © È ? ? ® ? di ta ?
Garudimimus brevipes È 0) IL il 0 ? È E E 0 1 Du 0 Di ?
Anserimimus planinychus È ? ? 2 ? fi È ? 2 ? Ci È ? d ?
Ornithomimus edmontonicus e 0 1° di Al 0 ? Il 1 0 e nl 0 nl Di
Struthiomimus altus R 0 1 1 1 0 ? 1 1 0 È ? 0 ? ?
Gallimimus bullatus 7, 0 1 1 Ji 0 ? 1 il 0 1 Il 0 dl 0
Falcarius utahensis Ù 0 gl ? ? 0 al (0) 0 0 P È (0) 1 ?
Beipiaosaurus inexpectus ? ? ? fe ? ; te E ? ? ? Ù ? ? ®,
Alxasaurus elesitaiensis ? Ù E È ? ? ? È ? ? È È; ? 2 2
Nothronychus mckinleyi ? ? ? ? È 0 1 0 ? ? ? ? ? ? ?
Erliansaurus bellamanus È, D ? ? G ? ? ? ? ? ? 2 ? ? ?
Nanshiungosaurus brevispinus ? ? ? ? ? ? ? E ? ? 7 ? ? ni ?
Neimongosaurus yangi ? 2? ? fa fi 3° fa
Segnosaurus galbiensis ? ? ? ? ? ? 2 ? 5) ? ? 2 ? ? ?
Erlikosaurus andrewsi ? 0 1 sl 0 ? ? 0 D Te. T R all ? 0
Therizinosaurus cheloniformis si ta È ? ? 2 Ei e R Ei n ? ? È sr
Alvarezsaurus calvoi ? ? ? ? e ? D ? 2? È ? ? a ? ?
Patagonykus puertai è? ®._ ? 2000
Mononykus olecranus R ? ? ? ? ? 0 0 ? ? ? ? R 3 ?
Shuvuuia deserti CIT 1 al al (0) 1 0 (0) 0 0 0 0 ) 0 ?
Incisivosaurus gauthieri e 0 al al 0 ? € ? 1 2 ? 0 È ? (e;
Protarchaeopteryx robusta ? ? ? ? e) 2 2, 2 5 2 ? ? ? 2 ?

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa /
Avimimus portentosus
Caudipteryx
Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis
Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi

Atrociraptor marshalli

Epidendrosaurus ningchengensis

Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

Characters

cd. IN

*V

*v

PRIVO TONNO Oo

*V

HW_P KH KR H HR _ KH bl HW KH Ww

“) “vd

*V

*V

> NIN <> INA ©

2A4S

220

Taxa y/ Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

1600170019 \01900200821 1022023 M5<4
0 (0) 0 1 0 0 0 1 0
0) 0 0 al 0 0) 0 Di 0
? Ci E fi € 0 0 1 0
È 2 5 2 2 2 0 ? 1
0 0 2 al 0 0 0 0 IL
Gi Fi ? ii 0 0 0 0 1
? 2 ? È ? È 0 È 0
2 2 2? ? ? ? 2 ? ?
1 2 ? ? 0 f: 0 1 0
E ? ? ? 0 0 0 L 1
? ? DI 7; 2 2 2 2 2
2 2 2 2 2 ? 2 2? ?
? 2 È ? 2 2 2? È e)
n ? ? 2 7 ? D ? ?
? ? 2? 9 e) e) È ? ?
C; fa ls S 0 0 0 0 n
? ? ? 2 2 ? 2 2 DI
È E 5 C 0 0 0 Il a
g E ot: È 0 D 0 1 ?
6 E 5 6 0 0 0 IL G
È ? È 2 0 0) 0 IL ta
1 0 È 6; 0 1 0 1 E
2 2 2 2 ? ? ? ? 2
E 5 t ‘dI 2 al 0 0 (0)
al C; 2 5 2 di 0 0 0
fs E ? A 2 il 0 0 0
? 2 2 e) e) ? ? ? 2
È 2 E 0 2 1 0 0 1
? D 2? p 2 ? ? ? 2
il 0 1 0 2 ol 0 0 db
al 0 1 0 2 1 0 0 1
1 0 1 0 Z al 0 0 1
0 1 2 il ? 2 d E E
P) 2 2 2 2 ? 2 ? ?
2 È e) 2 2 2 ? D 2
E E 2 0 1a ? È È Pe
2 2 2 ? 2 ? z ? 9
? È ? 2 2 2 È ? 2
? ? 2 2 ? E D) 2 )
? 2? ? 2? 2 È ? ? ?
i R E 0 1 (0) 0 1 al
È È È ? Gi 2 ft; p ?
? ? 2 2 ? 2? D 2 2
? 2 6 ? 2 ? E E %
il i 2 È ? ? 2? ? CP)
l 1 2 0 1 0 0 0 0
D 2 n 1 0 0 0 Di 0
2 2? 2 2 2 2 0 Dv) ?

DTA er A E

P_p

fi _

*v

0
0
)

*v

++ O

0
fi
Il
ti
sl
il
E
?

al
1

30

0
0
0

*

(3 SINNI ©,

*V

sini

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

18

5

1908200021
0 ? È
D 0 2
? G; ?
È È E
0 ? E;
È E 2
0 0 È
? |: 0
0 0 0
& 0 0
? ? E
È 0 0
È 0 0
fs n ?
0 0 0
? pi 1
0 1 al
È ni ?
3 P; ?
0 al È
n 1 al
0 ii È
È S È
2 ? 0
È ? Fr;
? C 0
E ta È
? ? 0
3 Z 0
1 2 0
È È ?
al Z 0
? ? D
al a 0
C z ?
0 ? ?
? E; ?
2 È: t
? È 6;
0 al 1
E 1 1
È al sd
se 1 1
. gi 1
E: I ti
? ci 1
E: C; È;

220230024
al o al
il I 0
Ci h fs
G 2 ?
fa 1 R
? ? a
? ? ?
iL il 1
i sl di
Jl 1 1
iL di al
al il il
il Il 1
2 È È;
0 1 0
0 gl 0
0 al O)
0 0 0
? Gi fo
di ? 0)
0 0 0
0 0 0
& ta È
0 0 0
? E ?
0 Di 0
0 0 fs
0 ? ?
0 E ?
0 0 0
? 2 2
0 0 0
5 ? È
0 0 0
È ? ?
0 È 0
0 Ci 0
È ? se
? ? v
0 1 0
0 Ci 0
? ? ?
0 1 1
0 0 0
0 È C4
il ? 0
? G iO

25082 60062.
È 5 0
? 0 Il
ti E È
? fo ?
1 al ?
Fi ? 3
G al 1
e 1 fi:
di dl Ji
il 1 al
il il 1
% 1 gi
G al pi
E Pl &
il E il
G fi È
1 0 i.
? È il
È; ba G
fi: 0 Di
gi Te 1
È: ? al
o È Hi
al 0 al
? Vi t;
2 0 IL
io 1; al
? VW E
È P 1
al 0

ts e È
st 0 Ii
© 0 aL
? 0 I
? 0 il
È 0 e
® f: e
? f di
? È; ?
È e 1
G f; È
ti 0 ta
€; & È;
? 0 fi
? i; È
? È, ?
i; fe È,

28

2986830
2 d;
0 0
È ?
È È
0 È
È È
È 0
Gi ?
0 0
2 0
0 0
0 È,
È 0
? È
fe 0
? 0
€ 0
? ®
È E
ll 0
0 0
0 È
Cs 5
? ?
e È
sl 0
dl 0
e È
is ?
I 0
P ?
al 0
di 0
iL 0
1 0
? 0
? ?
dl 0
? È
1 0
fi: v;
? È
? 0
d 0
? 3
S c;
? È

221

222

Taxa 7, Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

31320733. 034 350 36, (3)
0 0 0 dl al 1 0
0 0 0 0 0 1 0
0 1 0 0 f 0 ?
Di ? 2 ? ? È È
0 0 O, 0 0 0 0
0 0 0 0 0 0 0
0 2 6 6; 5 0 0
D ? n ? ? ? ?
0 al 0 dì Gn 0 0
1 È È 0 È di 0
2 D ? © 9 n 2
2 ? 2 ? ? 2? ?
? Co) È D D 2 ?
? e) 2? 5 ? ? È
È 2 p È ? 0 0
0 l P 1 R ? 0
? ? ? ? ? P) 2
0 1 d L È 0 0
ti 0 8 Ù 0 c 0
? 1 ? ? 2 D 0
PIO ERO ? "e, n G
0 0 0 IL R 0 0
? 5 D S 2 e) ?
0 È) 9 n ? ? 2?
0 ? ? ? ? D 0
0 ? ? Ti È PI ?
? È 9 D > D ?
0 ii 0 ? È 0 0
2) 5. ? ? ? ? ?
0 gi 2 Il al 0 0
0 ? 2 gl IL 0 0
0 1 È al 1 0 0
2 ? È ? D 2? ?
? > ? ? n ? ?
È 2 2 È E D D
S e) ? ? ? 2? ?
fe ? ? 7 ? e) 2
? 2? 7) ? 5 ? D
2 2 ? ? 2) ? ?
? 2 ? i) n P e)
0 ih 0 al ? i 0)
È o 2 È G) ? ?
9 5 ? ? È DI ?
? ? 2 ? ? z. 2?
2 2 ? P) ? ? ?
0 iL a st di 2; ?
1 1 0 di o 0 2
» 2 D 2 ? ? D

3839
al sl
sl 1
0 1
0 E;
0 1
il al
0 0
È ?
0 0
0 ol
2 0
2 G
È Ri
? 1
È 2
0 1
È ?
0 0
0 1
0 0
0 0
0 0
? Ci
0 0
2 0
0 0
© ?
Z 0
? ?
0 0
0 0
0 0
? G
P ?
2 ?
È ?
Ci Gi
Ts f
iv i,
© Si
0 0
bi; d
B P
e È
? ?
0 0
0 al
? ?

40

0
0
2

IS)

(2 RSM

*v)

41

0
0
1

DS)

43 44
0 1
0 di
0 2
È v
0 pi
0 1
? ?
E ?
0 (0)
0 0)
2 2
2 È
0 0
? io
E ?
? 0
Si; ;
Li 0
E; e
te 0
5; 0
1 0
? 2
? 0
0 0
0) )
? è
0 0
È È;
0 0
0 0
0 0
0) 0
® DÒ
? È
? f:
? È
? È;
f, ?
? e
0 0
E È
n E
f s
f; G
0 0
O) 0
? ?

+

I La

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

chi

?

321033 *434
Dl 1 a
1 1 1
È; ? F;
n 2 &
Cf; 2 2
F: E È;
DL gl 1
G ql 1
0 sh i
0 pi L
0 6 il
0 L 1
l 1 1
2 ? È

t; gf:
il 0 dl
fi È ?
? f; E
E b ?
1 n ?
? fi C:
È 0 f;
1a 5 ?
1 E È
Di È; È
il O) È
0 ? Ye
t ? 2
al 2 ia
? 0 gl
? È; E
1 0 al
b; È È
? 0 1
2 E; E;
1 0 ni
? È; se
Cs È 3,
fi d; È;
di 0 1
f; G p:
È È; È
1 0 1
di 0 1
? ? È
1 È E
f; 1 :

35

?

36 37
? il
0 0
È td;
9 ?
È fs
2 7:
fi ?
I 0
1 0
od. 0
ii 2
ÙI 0
1 0
? 0
7: ?
0 0
sa È
0) 0
t È
È; È
oc f
(o io
se È
È e
? È
1 0
I 0
È n
= i
al 0
2? ?
al 0
È E
1 0
È ?
di 0
0 0
f: ?
al O)
n O)
Le E
n 0
0 0
d E
? È
P È;
È; 2

38

ol aan feta be)

*v

39

40

41

e i

[oso ANI»)

223

224

Taxa 7 Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

v
LO)
v
v
v
> LO Oraria leo a

2 il ti 0 0

2. i L 0 0

2) ? 2 x ?

B ? 2 ? ? fp.

al 0 0 0 0 0
CIREZINRO 0 0 0 0

È 9 7, D ? ?

Fi ? D ? ? ?

p D 2 ? 9 9

) 9 ? È E È

a 23 E È ? O

? ? ? ?
? È ? 2? E ?
f de 0 0 P ?
0 6 0 1 0 È
? 2? ? ? ? e)
? 0 2 È ? 2
? ? È ? ? ?

O 2 DI ? D ?
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
? 9 ? 2 ? 0
? ? ? ? È ?
? ? ? ? È ?
2 È 2 ? ? È
? ? ? ? ? ?
? ? ? ? ? ?
? ? ? 2 ? È
? ? ? ? ? ?

? ? © n $ 2
? ? ? > n ?
? Z 2 ? ? 2?
e) ? 2? 2 ? DI
(0) 1 0 al 0 1
1 1 0 0 0 a

*V

e!

(le;

"Vv

54
0

(=:

*V

55 56 57
dui
0, Sola
go Lao
tar no
(ade ao
Ci ph
Dir Os
va:
#10 it
Oa PICO
A
(A
» Doge Ba
ca
». ene
pei feet
% Sa E
Geena
vi Aeon ha
di
marce e
"ATO
de LGS
Ra pa
DST RMMO
dre a
cid
age e
i
go Adi e
Di pei Ln 0
Oa UO SO
Ord Osasio
da
Di
Geni
FR PU
RM
Coal: let
a 9A
o oe
(o
dui
a a
Diana ai
O) OO
Dodo
co ge:

58 59 60
(0) 0 x
0 0 di
(0) : f;
È ? ?
0 0 1
0 0 1
Ts È £;
? È V;
0 1 1
0 0 ?
2 }; 2
2 T; y;
? C È
1a D ?
to È: È;
2 ? ;
? È; ?
fs 2 1
0 0 ?
t ? ?
di G Ù
E; ? 1
n ? 2
? ? ?
0 0 Vi
È ? E
2. fi; È;
0 0 ai
? ? È
0 0 f;
0 0 pi
0 0 1
0 0 >
fi V; ?
? È; fi
0 0 È
2 E ?
? ? d
C: 2 PR
a G 2?
0 0 D
? ? L:
È i f:
L, fi. È;
0 0 ?
0 fi; f.
a 0 È
? P ?

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

46
il

47
1

48

?

49 50 S5S1
È; 0 ie
0 C 0
K: % È
D È %
2 È È:
È t; È;
È È ca
0 0 6
0 0 0
7 0 0
E 0 n
? 0 0
6; 0 0
jé È 0
È f 0
al 0 Fa
È t G
1 fi 0
È of; È
fi 0 0
? È È
È 0 L
È : ta
È 5 0
6 È E
ia È 0
È C: 0
? sa ?
? & ?
0 si 0
? 2 È
0 1 0
z È: #
C di 8
& G ?
C: 1 0
? % È
? E fs
w 6; 0
0) 0 0)
C E s,
È G 3
È a 0
mi 0 Il
È V; (0)
È 0 1
? ? f

55 56
0 0
<“NIRO
PI e
FASE:
(COMODI
RAS?
n) 10
2?
gm
HALO
QLC
criuglalie.
GN
ua
cea
2?
1 v;
Dani AR
Rao
(O
A a
166071
col
SREEO
SATO.
<A,
7 3)
MES
sà. D
DO MEO
2A
0 0
alvnia
0 7
dn cia
A”
ue)
"1 ORE
aa
2?
Suap
2?
Data
200
2?
22
A.

Sul
1

60

229

226

Taxa /
Allosaurus fragilis
Sinraptor

Dilong paradoxus
Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii

Aniksosaurus darwini
Nedcolbertia justinhoffmani
Ngwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki

Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus

Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

Characters

CRISTIANO DAL SASSO & SIMONE MAGANUCO

61 62
urta
0
ca e)
; al
PACO
? e
D È
e) ?
=.
7 SD
e 7
? E
? SI
9 a
a
3000
yi
è ‘e
Sa
(BIO,
) be
a
sp) ?
) 2
0;
>) n
e) 92
? sp)
n tp)
be; 2
x ca
200
(O
pe) P)
e.
e) wp)
a
(I,
2 Di
pa
VIVI
3 PR
»
D Pi)
P. i:
dpi
dg 1%
PELI

63 64
0 0
CONO
sr
o
0. 0
» %
gr (a
FE.
BI
tao
uo
sia
x ea
c-
e
coli
an
i
- &
- 0
& (G
@. d
t 2
si Da
d 9
«e
*_
e. a
CO:
i la
0
» 3
= 4
ca
CARE
di
ca
CR
mia
Eli
ie _«J
—
7
ca
de a
ATE
pia

65

0
;;
3

Gi;

|)
0

66 67
d 16
COIN:
0 0
o.
o 0
0 0
> O:
vali
dt 0

0
pas:
| S
cale.
ca
es
cr 200
e;
0 0
‘0
0 0
o 0
o 0
Pat
o 2
o RUN
o 0
> ra
MINA ©
CA:
o 0
0 0
0 0
w 00
7
2 N
“20006 -
a
È la
E.
si.
PS.
(SAR
e
*
aa
e La

20
CAO

68

0
0
(0)

*v

Di

"

bt COSO

\ Ji = e]

dal

(0)
0
0

> i © 1.)

*V

(e ir © PI > |

1203
0 0
0

di C:
0 i
0 0
0 0
a E
gi ?
0 0
LL 0
? P
2 ?
fi. i;
fi È;
D ?
0 2
2 ?
0 2
0 Z
0 2
? 2
0 A
? d
fi 0
0 0
0 (0)
È: È.
0 0
9 È
0 0
0 0
(0) 0
0 È
? n
0 ?
? 2
fi ?
È C;
0 L:
? (0)
0 0)
È È
i; È
2 d
fi È;
0 0
0 0
0 ?

vw O Oo

fed e

*V

*

*V

9 Le

DS)

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa 14
Avimimus portentosus
Caudipteryx
Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis
Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi

Atrociraptor marshalli

Epidendrosaurus ningchengensis

Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

Characters

64

%

N_IN N

Vv

© @® ‘e. @ I © CS _ N

*v

"Vv

"Vv

Proto, (© NO e:

*v

"Vv

*v

o 0 O E E E EH HE EP HW _ HW HE

DI Ls] *v *v *v

"Vv

INDENT CSI INTESTATI

")

ele t-> @ © a: ©. Se vo er. ao

(hO)

71

pP_E O HW HO H E O HW

"v

(2:

*v OS TIAaAmRa

v

p_pP HP H' H' EF PP HW O O So o oo o o A

73

0

"O

P_KF HP H HE HF HEHHEHH||O I

*v o v

(©)

*v

74

75

0

220

228

Taxa / Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

0 0 0 0 0 0
0 ti 0 0 0 0
? È ? 0 0 0
? 2 2 ? 9 0
0 0 ch 0 0 0
0 0 di 0 0 0
? È 0 0 0 0
2) ? ? Fi ? 2
0 5 0 0 0 0
0 le 0 0 0 0
? ? © 7) ? 2)
2 ? È d ? ?
È; È ? ? ? 2?
? 2 ? ? È È
È 0 ? ? ? e)
0 Hi ? È P; 0
? 2? o ? n p
0 0 0 0 0 0
D 2: 0 0 0 0
© 9 2 0 © 0
0 G Si 0 È 0
0 0 0 0 ? 0
7) ? > ? ? ?
0 Ù 2, È i 1
0 ? ? 0 E 0
0 ? ? 0 ? Ji
? ? P È È ?
0 li 0 0) 0 il
? ? ei 2? ? L)
0 Gi 2 0 0 1
0 1 0 0 0 1
0 1 0 0 0 Il
? 5 È D E È
? D D ? p E
? 2? 9 ? 2? 7)
) ? 2? ? ? ?
n ? ? 5 ? 2
? D ? ? ? ?
? ? È 2 ? È
0 Fr: 0 0 0 e
0 1 0 0 0 di
? 2 ? ? 2 ?
? ? ? P È ?
x, D ? ? ? ?
2 p ? ? ? 9
Ola iL 0 0 0
0 0 ? sl Li 0
D ? ? È 5 0

OG. + (ST... 0

(esi = |

N_N N N

85

0
0
0

*v

DINA

*“v

P_H _ KR FF W UW W

86

p_H HH HE HW

87

O > o Sto

88

0 P_pPo_ HW

H_H HP

(‘uo SORDI «> AGI > FIRE «©.

89

0
0)
0

*

di

Lx E «> ENTI «©

90
hi
1

?

ri

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

*v

10. GOTTO TO>oRS>BOoOTOSo Re ro MR

pPo_H

*V

BS)

O_O P_E_ POOH Ho_H_ HE HO _H

DPO) I “Vv

*v

(@). fof.(°} logeft el de ietg>- ris, Le:

O)

*v

PIETER TNCSTN

‘Vv

229

230

Taxa / Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

91 92 93 94 ‘95 96 97
0 0 gL 0 0 0 0
0 0 al 0 0 0 0
1 d È È 0 0 È
si È È È 0 0) ?
al 0) nl 0 0 0 0
1 0 Di 0 0 0 0
0 o) 2 ? ? ? p
e È; E Ci 0 0 0
0) E È S 0 0 0
di 0 ? È Ù L; >
2? d ? 2 0 2? O
? 2 ? ? 2 ? ?
È ? o 0 9 ?
(È ? 2? ? ? 2 ?
? ? ? ? D 2? ?
0) 0 0 Ei fa E È
2 2? ? ? ? 2 2
0 0 0 ? 0 0 0
0) 0 È o È 0 5
0 0 Gi È 0) 0 E
0) 0 È n 0 0 ?
0 0 gl 1 0) 0 G
2 > ? ? ? 2 ©
È 0) 2 È È 0 0
1 D 2 2 Al 0 0)
? ? 2 ? ? 2 È,
? 2 ? ? 0 0
E; Ss 0 1 1 0 0
? ? 2 ? ® ? ?
È 0 0 Ji ? 0 0
E 0 0 il ? 1 0
E 0 0 È 1 0 0
n È 3 fa 1 0 È
? P 9 ? ? ? 2
È È 2 2 ? ? ;
fi 6 È È 0 0 0)
7) 2 D 7 ? ? n
7, 2 2 2? 7) ? ?
? 1 ? È Al 0 0
? ? 5 ? 2? ? ?
2 ? ? , 2 e) 7)
i; vi ? ? D ? ?
E, 2 ? ? 1 DI ?
? 5; ? ? ? © ?
2. 3. z 2 ? 0 al
© È 0 1 il 1 x
2 ? 2) 2 ? 2 p
2 n ? e ? 0 0

98

0
0
0

Ti

*v

99 100 101 102 103 104 105

LISTE: Pia

"o

(e ME ceo AI > MINA ce PO 0 TC AE cn NT I a = I

oa is;

*V

0

(Me)

*“v

0

reo eloeio

*v

4

9 HR E E HE

0

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

*v

*v

> MR)

99 100 101 102 103 104 105

FSE:

*v

I

2

O)

2

0

R

(0)

2

20

2392. CRISTIANO DAL SASSO & SIMONE MAGANUCO

Taxa / Characters 106 107 108 109 110 111 112 113114115 116 11/115 Li94812z0
Allosaurus fragilis 0 0 0 0 0 0 0 0 0 il È 0 si 0 0
Sinraptor 0 0 0 0 (0) 0 0 0 0 ii È (0) ? ? ?
Dilong paradoxus 0 ? d; C ? ? 6; È 2? 1 ? fa ? È 0
Eotyrannus lengi ? È G 1 È. E f: 2 fi E ? ? ? C; ?
Tyrannosaurus rex 1 0 0 0 0 0 0 al 0 si ? 0 0 0 0
Gorgosaurus libratus gi 0 0 (0) 0 0 0 1 0 gi Vi 0 0 0 0
Tanycolagreus topwilsoni 0 0 0 ? È ? G (e & È È 0 È ? 0
Coelurus fragilis 0 0 0 ? È È ? ? 5 ? E; 1 0 0 4
Ornitholestes hermanni 0 0 0 z i) ? Al 0 0 ? P al 0 si 0
Guanlong wucaii 0 0 0 ? 0 2 ? (0) 1 p ? È ? Z 0
Aniksosaurus darwini ? ? ? ? fi E; dI ? ® ? Di 7 0 A ?
Nedcolbertia justinhoffmani dl Gi ? v ? a 0 0 io fi fe 0 f; iL *
Nqwebasaurus thwazi ? ? F; 6 so ? s È Li ? 2 CARNE fe E
Santanaraptor placidus i è ie ? ® ? n È È È È ? È ? E
Orkoraptor burkei Do È ta ? È : ? ? ? 5 G È 0 ? H
Juravenator starki Z 0 0 2 ? 7 MOI 0 0 0 ? 0 È pi
Mirischia asymmetrica fs ? ? ? & 0) RO, G ? Fa 2 % ? ?
Compsognathus longipes 0 0 0 DI 0 P % ? (0) 0 0 te (0) ? 2
Sinocalliopteryx gigas 0) 0 0 to ia ? & ? ? 0 0 ? 0 ? 2
Huaxiagnathus orientalis E 0 0 2 p d; ? ? E 0 0 0 0 ta 2
Sinosauropteryx prima 0 R 0 2 0 ? ? ? ? 1 5 7 1 i 2
Scipionyx samniticus d 0 P 2 0 O ? R 0 CARO f. È ?
Deinocheirus mirificus 2 ; ? a n ? ? 6 fi: ? 4 C: G ? n
Harpymimus okladnikovi 0 0 0 (0) dh B di 0 0 0 0 ie (0) 0 (0)
Pelecanimimus polyodon È 0 n 5 È fi io K ? ? È f f Ve ?
Shenzhousaurus orientalis 0 ? E d 0 Dn ? fe 0 0 0 (0) 0 L Le
Archaeornithomimus asiaticus 0 0 0 0 1 ? Il 0 0 0 0 0 0 0 0
Garudimimus brevipes 0 ? 0 (0) 1 È R 0 ? p t: ? fe 9 p
Anserimimus planinychus E È 5 D È ? ? 6 9 d; È ? & ? ti,
Ornithomimus edmontonicus 0 0 0 0 al 0 1 0 0 0 0 0 0 0 0
Struthiomimus altus 0 0 0 0 iL 0 al 0 0 0 0 0 0 0 0
Gallimimus bullatus 0 0 0 0 il ? si 0 0 0 0) 0 0 0 0
Falcarius utahensis 1 E; 0 0 0 ? J al 0 0 1 1 1 0 2
Beipiaosaurus inexpectus ? ? ? ? in ? È ? D ? ? ? ? ? n
Alxasaurus elesitaiensis 0 0 0 0 0 ? di 8 0 0 pl ? È 0 Z
Nothronychus mckinleyi 0 0 0 0 1 Ci 2 di i ? ? 1 ? pi: ti;
Erliansaurus bellamanus 0 ? ? n iO ti P E E ? ? ? 0 3 2
Nanshiungosaurus brevispinus 0 0 ? ? (0) f. ? ? (0) ? ts Ù; ? È È
Neimongosaurus yangi 0 È 0 È al ? È 2 2? 0 al 0 0 0 2
Segnosaurus galbiensis ? ? ? ? ? È ts ? 0 ? C; Ù; È È ?
Erlikosaurus andrewsi ? ? ? ? ? ? 9 ? ? ? va ? ? ? ?
Therizinosaurus cheloniformis 5 e e Le 2 ? fe E L: ? e ? ? C ?
Alvarezsaurus calvoi n x ? ? ? fi Z ? il ? Ve Zi A 1 z
Patagonykus puertai ? E. ? ? p 2 ? ? il ? 7. 2 ? ? ?
Mononykus olecranus (0) 0 gl di È; 2 È ? 1 S ? 5, ? fo td
Shuvuuia deserti 2? 0 9 2 1 2 2 0 1 ? 2? o) 0 1 2
Incisivosaurus gauthieri ? ? ? ? ? ? ? ? ? ? ? 2 ? ? ?

Protarchaeopteryx robusta Fi Ci 2 d 2 2 9 ? È 0 il ? ti; Ci D

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

Mo 7 Sio) alime) slalo) sale ail) sile} abi also) SUS Gli) Si; Glelef ale

miti

HY O Ww

&
È
al

u

0

?

0

di

0

PR

0

È

"

*v

IENE NTEN ZE TOTO TEO

BS)

1
Hi
?
G

1

0

®

0

0
(0)

È

NN

*v

233

234 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Taxa / Characters 121 122 123 124 125 126 127 12801290130 131 1320133 054280
Allosaurus fragilis 0 0 0 0 0 0) al 2 Hi È 2 0 0 0 0
Sinraptor ? ? ? so ? ? Ci al 2 0 0 2 0 20000
Dilong paradoxus © 0 & 0 2 1 ? 0 e È 2 2 D i (0)
Eotyrannus lengi f; a È È ? d; Ga P ? ? 5 ? 0 È 0
Tyrannosaurus rex 0 0 0 0 2) ? Il ? ? Pd 4 0 0 0 0
Gorgosaurus libratus 0 0 00 @ d 0 da © a GA
Tanycolagreus topwilsoni ? ? È ? ? ? d dl È F; ? È 0 0 0
Coelurus fragilis È 2 E È ? È È È ? 2 ? ? 0 ? 0
Ornitholestes hermanni È È 1 ? ? 2 È È ? jo fe ? E ? E;
Guanlong wucaii e 0 ? ? 0 E: È ? P Ci ? e 0 0 0
Aniksosaurus darwini D; È 2 P 7: te ? ? ? ? 6 ? fs È i
Nedcolbertia justinhoffmani ? 0 ? È ? f: ? ? & ? È 2 R È 0
Nqwebasaurus thwazi È e n È ? Li ? e È P ci RATIO (0) 0
Santanaraptor placidus ? È R ? 2 È È È È È P E È ? si
Orkoraptor burkei È 0 D & ? È i: o H w ? ? ? ? di
Juravenator starki 0 0 CANTORI AEO 0 0) (0) 2 0 0 te 0 0 s
Mirischia asymmetrica È; È ? 3 ? 2) ? ? ? ? ? ci ? ? ?
Compsognathus longipes 0 (0) 0 0 0 0 pi te È ? ? ? 0 0 3
Sinocalliopteryx gigas 0 0 0 0 0 (0) 0 ? ? ? ? 0 0 0 0
Huaxiagnathus orientalis ONELO 0 (0) 0 0 0 ? È ? ? 0 0 0 0
Sinosauropteryx prima 0 0 0 0 0 (0) (0) 0 ? ? ? E 0) 0 Di
Scipionyx samniticus 7: 0 ? 0 0 0 di È n È Ci 1 (0) 0 g
Deinocheirus mirificus dI ? D Ta È ? ? E ? 2 C p ? R IL
Harpymimus okladnikovi 0 0 P 1 P Pe ? ? ? ? ? p 0 ? 0
Pelecanimimus polyodon 5 3 ? ti PR È e 0 1 (0) 0 È ; n (0)
Shenzhousaurus orientalis ? 0 ? È ? E ? ? ? ? ta Li ? ? 2
Archaeornithomimus asiaticus e 0 E P ? ? È ? ? 7; ? ? 0 59 0
Garudimimus brevipes 2 r £1 oa» 0 è a RR i

Anserimimus planinychus È ?
Ornithomimus edmontonicus t:
Struthiomimus altus
Gallimimus bullatus

Falcarius utahensis ?

ni P.
d il iL
1 ni n
1 1 1
1 9 2) o; 2 ? ie) n ? o 0 0
Beipiaosaurus inexpectus al 2 f; E Vi fa ? f: | ? ti 0 fi È 2°
Alxasaurus elesitaiensis ? 0 ? al Di ? 2 Ta ? 7, ? 9 È; ti ?
Nothronychus mckinleyi 1 E ? ? È ? & e L: ? ? ? (0) ? 0

Erliansaurus bellamanus Di D; ? È ? E VI G È ? R di fe ? v:
Nanshiungosaurus brevispinus ? È È ? p ? È L: R È fo E ? ? si
Neimongosaurus yangi 1 0 ? 1 R Le ? ? ? E ? 0 0 0 al
Segnosaurus galbiensis È ? ? ? ? ? 2 ? ? ? ? ? (0) ? 1
Erlikosaurus andrewsi E; È ? pf: È {; 2 i PD di Vs La 1) 2 i;
Therizinosaurus cheloniformis fo ? 7 f v) D e È ? C; (i ? (0) 0 1
Alvarezsaurus calvoi ? ? ? 7 Ed ? ? ? ta ? ? ? 0 ? 0
Patagonykus puertai vt *. £ * © @&@ @ &re de e
Mononykus olecranus ? ? ? 1 L; È ? Il 0 0 0 di 0 (0) 0
Shuvuuia deserti O gd 0 1 0 0 È 1 0 0 0 va (0) (0) 0
Incisivosaurus gauthieri ? ? 2 ? ? È ? ti 0, ? ? ? 2 p 3

Protarchaeopteryx robusta Al 0 & 7; 7) P o 0 P Pi vi 0 È; ? ri

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa /
Avimimus portentosus
Caudipteryx
Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis
Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi

Atrociraptor marshalli

Epidendrosaurus ningchengensis

Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

Characters

?

al

f;

SIINO

*V

?

1

?

È
1
£
in

?

*v

?

0

2

To

H_ pP_ KE HW EH

*v

oO H oO HW

O O H f_ H_ H_F HW HE HE

"Vv

Ò

*v

il
dl
sl

21001220812 3812491258126812:/12.88129 8130813 19132713381348135

1
0
0

[e I 00 RI US 0 TORE e DID > (I co e co.

*v

fe)

236

Taxa 4 Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

136 137 138 139 140 141 142 143 144 145 146 147 148 149 150

0 0 0 0 0 0
? ? 0 0 C: È
2 0 bi il 0 ?
? 0 Gs 0 0 0
0 0 0 0 pi 0
0 0 0 0 al 0
0) 0 0 0 0 0
fi: on 0 È (0) 0
V E 6 f: 0 ?
Te 6 G 0 0 C
? se 2? fe 0 0
È 9, 3. %. È È
0 0 0 0 0 a
D ? ? 7; ? n)
? ? ? ? © ?
? i 0 0 0 ?
5) n ? » ? D
0 0 0 0 0 0
0 G 0 0 0 È
0 0 0 0 0 NO
0 0 È; 0 0 f
0 Ss 0 dl 0 0
vi 0 0 0 0 0
2, 0 É 0 0 0
Z ? È ? 2 ?
D Di 2? >, E ?
Z 0 di È 2 0
p ? 2) ? ? ?
SE © di di 2 2
2 0 di 0 2 0
2 0 ti 0 2 0
2 0 LL 0 2 0
E dl 0 0 0 1
0 ? ? ? ? È
hi d sa 0 0 0
? ? ? ? 0 a
bs P 5 0 0 te
? ? » ? ? ?
0 0 0 0 0 0
0 f di Vs 0 0
? ? ? È 0 ?
0 0 0 fe 0 ?
? 0 ? ? ? ?
2 0 Fi; È 5 0
2 0 0 0 3 0
2°, y 0 | 0 3 0
DI D ? 2 n p
1 I 2 ? È È

al

0

(o) ekec)y Mico) Meo Mio) ee leo)

*v

*"v

0

O Tora:

0
?
ds
ti
?

0

di
?
È
Ci

%

0
Il

*"v

0

[ex I N N © A © N > N

*v

1

Gi

0

*V

N _N

0

O. O È» è © è. O

=> POOR => TU c» SAAS —- VOI <> ORI > DR — E e = |

*

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

136 137 138 139 140 141 142 143 144 145 146 147 148 149 150

?
1
0

*v

*v

HP Po _H HP _ HEET HtK@©«H

*v

Ei

*v

W W W HE W E HE

*V

Pò. P._&Ooo o o, O

IO SOLI

*"

fan

DS)

pP_RPo_ LE _HKE EHE HW

"

*v

PO loro TS9Ttortol o iii ie

"

o

‘v

*

Corna Ra RO TOO De

*V

0

È

1

*v

*v

o o

*v

Too Ordo Dia

*V

(©)

*V

(©)

"0

ol ele el omo)

"

?
0
0

fi

«È
D
0
?

?

E
?

ti;

HP_Ho HP _ HE _ KH Hp

*v

O. H «HW KH KH H_ O O O

ST NT TEST

e
0
®
ta

?

*v

*

Coi mac) ie Gi Mico af o Sai co Tg ee RR > AR >]

5

Da

238

Taxa / Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

151 152 153 154 155 156 157 158 159 160 161 162 163 164 165

0 il 0 L 0 0
0 Hi fi gi 0 0
0 0 0 0 (0) 0
0 0 z te ti E;
0 0 0 al 0 0
0 0 0 al 0 0
0 LL 0 È 6 0)
5) ? 2 ? ? ?

0 0 0
0 0 1 0 010)

1 0 ? ? A

dl 0 G; ? f;
? ? ? ? D ?
? ? D 7) di ?
0 0 0 0 0 0
é v; » 1 © 0
0 0 0 0 0 0
0 0 0 al 0 0
0 0 0 0 0 0
0 1 0 0 0 0
0 0 0 al 0 0
0 0 0 2 ? ?
z 0 0 Il 0 2
Z 0 0 È t; È
2 0 0 di 0 0
i 0 0 ? ? È
? ? 3 sl 0 0
1 0 0 ti ? È
1 0 0 1 0 0
1 0 0 1 0 0
1 0 0 IL 0 0
0 0 0 il 0 0
0 0 È al 0 0
0 0 0 1 0 0
D ? ? ? ? O
0 ip 0 ? e) 2
Cs ? ? al Y, 0
? 5 ? 1 p È
È ? ? il 0 0
D ” 5 7) © ?
0 sa 0 7: ? ?
TL 2 3 ? 0 0
1 1 0 ,; È È
sf sl 0 0 PI ?
nt 0 0) 0 5;
? ? È ? ? È

0
0

f;

0
0
di
fi
0
0
?

o

0
E
?
?
0
0

2

0
0
?
?
0
0

n

ll
1
?
fi

a
dh

0
0
n
f:
0
0
?

È

al
0

0
0
6

2

iP)

0
Si
fs
3

0
0

*

Lc NO >

0
0)
0
R
0
0
5
È
0
0

È;

*V

"= =)

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

1552815315481 550156815 / 15 BRL 59RL60RL0121602081638L604#8105

?
0
0
il
0

2

Sto ero To ao rrarana

")

oe =]

oO e... Lo. QD

© © E è-:i_ la >îea_Oo E. _ _S o_o

[en > e)

*v

*v

Para ctaortoì Co DID

di IO ot ATOMI

*v

QUO

0
1
i

Dore a. a S LD

0
0
0)

?

O)
0

2
Z
2

INN ISS CUNTIIST INC IINTN

*v *

*v

al
0

?

p_ Po _H Po _H_H_ HEHE "HW

"v "o

*v

1
«e
È
?

E,

t
?
&
È
0
?
0
&

0
0

[<> i >

0

?

0

[el =)

(Rs)

1
1
gi

0
0

?

299

240

Taxa / Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

166 167 168 169 170 171 172 173 174 175 176 177 178 179 180

0 1 0 Z 0 0
0 1 0 2 0 0
3 il 0 3 È 0
È E G ss È t;
0 1 0 3 1 0
0 gi 0 3 1 0
i? 5 A SIE ?
i 7; Z 5 ? È;
0 li 0 se 0 0
0 di 0 3. 1 ORO
? n ? ? 0 ?
to P 0 2 0 P.
? ? 0 1 È ?
0 ili 0 £ 0 ?
& ? ? ? ? D
(0) al 0 ? È; A,
0 1 0 3 0 SALO
0 al 0 3 v di
il si 0 2 E ORO
0 il 0 2 È 0
0 dk 0 2 ? 1
0 l ONE R3 e 0
2: ? ? ? ? ?
0 it ? D ? ?
? ? D È ? 9
0 Ùi 0 al ? 0
1 dì 0 ? 1 Va
? ? ? 2 ? ?
? ? ? ? ? 2
Il IL 0 2 1 0
L ol 0 2 1 0
ll dl 0 2 il 0
2 2 0 al 0 di
? ? ? 2 ? ?
0 2 e e 0
0 2 Li 2 0 o
? ? ? ? 2? ?
0 2 0 ? 0 E
? ? ? ? ? ?
0 2 di 0 È 0
? ? D È È; ?
? ? ? ? ? È
? ? ? ? ? ?
) ? 2 2 ? 2?
0 0 È È o f;
0) 0 ? 0 0 0

0
0
©,

?

a
0

0
0
al

0
0
?

?

SMOi MO MOSTO

0
0

Ci

0)

0

0
0
0)

0

*v

0

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

TOCRL670 ‘6881698170 017.10817201573817/481:/5 01/601 1011891/98180

?

2

È;

*v

*v

STEPS NTNZEN

‘Vv *v

*v

*v

H KE QQ H_ HE _ HLkH_E «O

*V

0
?

î;

HP_HPo _HgoOHWw

0
ti;

6

1
1
1

*v

HP _pP _H 4HE_ KH KH EE _ N

H_H vUH ww

0
0
0

Il
?
1
5

0

0
1
l

*v

O)

O O E HE HE H HP O_HP

CENTO)

*"V

0
0
0

0
?
2
È

È

241

242

Taxa / Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

181 182 183 184 185 186 187 188 189 190 191 192 193 194 195

0 0 0 0 1 1
0 0 0 0 il 1
? 0 ? ? ® ?
7; ? Dì ? » È
0 0 () 0 A 0
0 0) 0 0 pi 0
0 0 0 0 sl 0
0 0 0 0 il 0
(0) 0 0 1 ? fe
0 0 L 0 1 :
0 0 0 0 sl 0)
0 0 È 0 0 0
0 C; 0 E 0 D
0 0 0 0 0 n
? ? ? D 2? ?
Oli 8 0 2 ?
0 0 0 0 sl 0
0 0 R 0 6 Fi
? ? e, È 2 2?
0 È & È 0) n
0 0 A 0 0 ?
0 0 0 al 0 ?
? D 2 2? ? ?
? 2? 2? 2 7) 2
P) ? 2) ? 2 2
0 0 0 0 il e
0 0 0 0 1 ?
0 0 0 0 il 0
? D D 2 ? ?
(0) 0 0 0 1 ii
0 0 0 0 L l
0 0 0 0 di 1
0 0 0 0 al 0
0 (0) È; 0 n, ®
0 ? 2 2? ? n
? d ? 0 2 e)
b P 0 È: Ci 0
2? ? ? 2 È ?
0 È 0 0 0 0
r ? n ? ? 9
? ? ? 2? ? D
? 2 ? 2 ? ?
1 ? ) 0 È 2
al nl d 1 ? 0
o 1 0 1 0 0
LR 0 1 0 0
È ? D 7. ? ?
1 R g Li 0 D)

0
0

SO re rle Gallo rta, holy

*v "

*V

0

(27 it o)

"vv

0

fa aio to o.

0

*

DO.

0

al

0

p_p HE HW

0

(© i © © — )

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa ( Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

181 182 183 184 185 186 187 188 189 190 191 192 193 194 195

*v

P_HE HE HE Ro _HW HR _HSO ev

Po_Ho vo Ho sv

iL

5

0

?

*v

Uma IG ROERO

Risto ne

(21

*V

H_ PE Po _E Ro _H HR Ho my

vd Ho HR dd RO

pi

ee)

USI

SO)

p_HE pf _R_E EHEH < NW

*)

0
sl
1

*v

(©) Go © 5 DI

"0

55)

*v

(=)

©

(S|

Di

(<© TBE, |

?
È
0

di

0

E

0

*v

> Lena te MEN SRL, SO DT

*v

sl
?
1
&

al

0

È

0

(0)
0
0

1
1
1

1
0
0

243

244

Taxa / Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

196 197 198 199 200 201 202 203 204 205 206 207 208 209 210

0 0 0 0 0 0
0 0 0 0 0 0
2 ? È 1 0 0
0 0 2 ? ? ?
0 0 0 0 Z 0
0 0 0 0 Z 0
0 0) 0 0 0 0
0 0 G 0 fe C;
(0) 0 0 0 0 È
È 0 È; E 0 0
0 0 0 0 0 RO
0 0 0) 0 al 0
0 0 è) 0 0 E;
fe 5 (a i 0 fi
? ? PI ? ? 2
0 0 %; ? 0 0
2 ? D ? ? p)
0 0 0 0 0 0
0 0 de c: 0 0
0 0 0 0 0) 0
0 0 0 0 0 0
2 ? 2? ? ? ?
? 2 È 2) ? È

2? ? D 2 ? ?
2 DI , ? n 2?
0 0 0 ) il È
0) 0 0 0 1 Fi
0 0 0 0 2 fi;
0 0 0 0 2 0)
0 0 0 0 2 0
0 0 0 0 2 0
0 0 0 0 0 0
0 0 fi; hi fo fi;
f; 0 0 0 0 0
? 7 0 n) È ?
2 2 2? e) P) D
2 2 2? ? ? 2
0 0 0 0 0 ?
9 0 0 0 0 f;
0 È; 0 0 È 0
2? 2 2 ? ? 2
0 0 0) 0 0 0)
@ 0 tf; E 0 È
0 0 0) 0 2 0
0) 0 0 0 2 0
2? ? 2 2 ? ?
& 0 G ? D 0

0
0
?
?
0
0

6

0

(e es)

*v

0

(0)

*v

0

0

0
al

?

DS)

0
0
2
E;

0
0

0

(<> Jaggmei «— i < TOi,. = OA — RN I —

L= = IR = co pe <> i = I De)

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa vi Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

196 197 198 199 200 201 202 203 204 205 206 207 208 209 210

*v

so RAIN co FAIR 0 VG =" RN co i «> ST © I)

*v

©

*v (©) ©} *V

*V

(SOA

*v

ND HE La WWW HE_ O

*v

*v

[w») PITOCVOl> oO Tao Loano

"vv

e (i;

l'as Mei: <> INR <> (AA © ©

USI

H_ Ho HE

2
2

?

fia. © © N

(IO Oi

ear ep Proteo t

“Vv

FO ES EAT a

*v

(> NERE > MA Mc UT Cc» E — TARE: «> TOI = ©)

*v

fs

0

?

0
0

*v

?
0
?
0
0

0
0

2

*v

aio. © TOTO TOnhOoT>9TOoOTo

*v

p_tE HE HW

*v

SIRSEOTROTRO To OD

*v

0
0
0

*v

*v

*v

SISTO To o 9 T-T-S CS

*)

0
0

È;

*v

(LORO OTO ROTIOS

[e] * *v

*

Ù

(> RAMI > 0 ae © si)

*V

coso DI RI ct ie) «YI 0 DI pl 10 N <> N <> DO <> I

245

246 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Taxa / Characters 21102120213 82140215021 60821:7082:1/882:1982208221822208223 0824 0d
Allosaurus fragilis 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0
Sinraptor È 0 0 0 ? 0 0 fi 0 0 0 0 0 O) ?
Dilong paradoxus È 0 ? ® C 2 0 ? 0 P ? ? ? 0 È;
Eotyrannus lengi Ti 0 ? ? 6 ? 0 0 2? 9 da P ? ? 0
Tyrannosaurus rex 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0
Gorgosaurus libratus 0 0 0 0 0 0 0 0 0 0 ? 0 1 0 0
Tanycolagreus topwilsoni 0 ? È (a dI ? ? 0 0 0 È ? ? ? 0
Coelurus fragilis 2 ? R ? 0 i? 0 ? fs 0 È È ? O) 0
Ornitholestes hermanni 0 0 al 0 ? 0 0 ® 0 2, 0 0 €; 0 0
Guanlong wucaii 0 0 0 0 È 0 0 0 0 R fe 0 ? 0 2)
Aniksosaurus darwini 0 2 ic IL 0 f; È 5 i 3 7; ? ? ? 0
Nedcolbertia justinhoffmani È ? d s ? ? 9 e È F; ? È ? G ?
Nqwebasaurus thwazi 0 ? È; ? ? È ? ie E È G e ? È; 0
Santanaraptor placidus fa ? O ? e P 2) di ? È t; ? E 6 ?
Orkoraptor burkei 9 È ? È; ? Re COS ? Aa È; C È: ? ?
Juravenator starki 0 0 ? 0 f 5; 0 0 ts 0) È fi È 0 0
Mirischia asymmetrica R È 2 ? e ? n: ? È Sa fo È: ? È. 3
Compsognathus longipes 0 0 È 0 Lo D; 0 0 0 È di Va ? 0 0
Sinocalliopteryx gigas 0 0 ? E: È ? 0 0 0 È ? ia ip 0 ?
Huaxiagnathus orientalis 0 0 ? È È ? 0 0 O) 0 È ? ? 0 ?
Sinosauropteryx prima 0 0 0 R 0 0 0 0 0 0 & 1 ? 0 ;
Scipionyx samniticus 0 0 ol (0) 0 0 0 0 ? 0 È e ia 0 0
Deinocheirus mirificus ? it; D Fa E ? ? 0 o 0 2 Pe ? ? 0
Harpymimus okladnikovi 1 1 0 ? ? ? al ? 0 0 ? ? È (0) 0
Pelecanimimus polyodon si 0 È a P 2 rl ? ih To e ? È 0 ?
Shenzhousaurus orientalis ? aL 6 0 0 I al ? ? ? ? ? ? 0 È
Archaeornithomimus asiaticus fl ? ? IL 0 F; ? di Di di fa E to 2 0
Garudimimus brevipes fi 5: ll 1 0 È & il 1 ? 0 0 al 0 ?
Anserimimus planinychus Il È od 1 0 ? ? 1 ? i Ù ? ? ? ?
Ornithomimus edmontonicus 1 Il 1 I 0 1 ? 1 1 1 0 0 4 0 (0)
Struthiomimus altus il 1 1 il 0 1 6 L 1 1 0 0 Mi 0 0
Gallimimus bullatus sl DÌ gl il 0 ah ? il 1 1 0 0 il 0 0
Falcarius utahensis ? ? ? 1 0 ts 0 0 Ca 0 ? 0 di 1 IL
Beipiaosaurus inexpectus ? È 6 ? ? ? 0 © E ? ? ? ? 1 1
Alxasaurus elesitaiensis ? i) Fo Il R È 0 0 Vi 0 ? 2 ? di È
Nothronychus mckinleyi ? D f ? È ? 0 È 2 ? P È 0 È £
Erliansaurus bellamanus ? ? ? ? Di g ta È 0) ? ? 2 ? ? 1
Nanshiungosaurus brevispinus 6 e tf: ? Ci fi se 2 di ? è iO È P È
Neimongosaurus yangi È; 0 È 1 a ? ? 0 P 0 È È 2 2 a
Segnosaurus galbiensis ? ? È ll È ? 0 iP 0 ? ? ? f 1 pi
Erlikosaurus andrewsi È 0 1 D È 0 0 % 0 ? 0 0 0 1 L
Therizinosaurus cheloniformis ? ? ? e ? È G 0 ? 0 ? n È È; 1
Alvarezsaurus calvoi te f; ? 1 È È; e 0 ? ? p ? ? ? ?
Patagonykus puertai e 2 ? ui 2 ? ci) 0 fe 9? ? ? de 2 0
Mononykus olecranus 0 ? ? È. 1 ? E 0 2 0 F; ? D 2 0
Shuvuuia deserti 0.0 0 sl l 0 0 (0) 0 0 0 L 0 0 0
Incisivosaurus gauthieri ? 0 0 ? ? 0 0 ? 0 ? a (0) 0 0 ?
Protarchaeopteryx robusta 1 0 2 ; ? ? ? e 0 ? ? ? ? 0 0

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa /
Avimimus portentosus
Caudipteryx
Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis
Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi

Atrociraptor marshalli

Epidendrosaurus ningchengensis

Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

Characters

*V

pP_KH KH HE HE HE HE HE_N

*v

*V

DORSO O O LEI

6) “Vv Dee)

ORSO LELLO LOST TO lag

*v

0
?

?

ol
1
gl

vv

p_po_pR_ H HR _H HE HW

* *

*v

?
ùl
1

0
0

fi

fi

È

6)

*v

Piet (@D esi Moebrier ret uo)

*v

(p)

(hO)

STtDLOd Oo TOTO TOTa:

?

0
0

(o aio (© a Me Ae Ao a)

"Vv

0
0

?

*v STO io SISTTOTOMNOSOTOoTo Teo

ee;

0
0
0

0
?

È;

0
2

2?

0
d;
È:
f;
0
fo

0

?

0

?

(© PRA > RE ="; > QUIIIOY «> MIE «> RIN o A IO o e > co AC]

*V D Cd *V

Oort

2108 2.12082:03 821/482 1582.1/6821:782:1/682:1982208221082220822372248225

0

Fe

0

d

*v

Dr da e SS I

*v

247

248 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Taxa / Characters 226 22:71 228° 2290230 23102320233923492357236423:/08238902398240
Allosaurus fragilis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Sinraptor di 0 0 0 0 0 0 0 0 0 0 0 0 0 (0)
Dilong paradoxus vi ? 0 E 0 0 0 0 OZ (0) 0 t) 3
Eotyrannus lengi 0 G G E 0 0 ? 5 0 f; 2 0 0 D È
Tyrannosaurus rex 0 Il 0 (0) 0 0 0 0 (0) 2 0 0 0 di 0
Gorgosaurus libratus 0 Il 0 0 0 0 0 0 0 2 0 0 0 1 0
Tanycolagreus topwilsoni 0 È È ? ? f ? ? Gi È i 0 ? È ?
Coelurus fragilis 0 ? 2 z È ? 3 ? 7) È ? È) Pi; ? 3
Ornitholestes hermanni 0 1 0 0 0 0 al 0 0 1 1 0 0 1 0
Guanlong wucaii 0 Il E 0 0 0 Il 0 0 2 0 1 0 1 0
Aniksosaurus darwini 0 5 Ci 0 B ? fs È ? ? È fi; ? Fi È
Nedcolbertia justinhoffmani 0 0 ? È È 2 P te È ? È E ? w ?
Nqwebasaurus thwazi Gi A È Ù ? f; ? È ? ta Li ? 2 vd; f
Santanaraptor placidus sn 0 Di ? È È; È 5 ? e fo sE K: p ?
Orkoraptor burkei ? ? 0 & 0 È 9 & A ? i: ? Yi; ? fe
Juravenator starki 0 3 0 0 0 0 0 di 0) 0 gl 1 0 0 4
Mirischia asymmetrica 2 0 to 0 ? ? Li ? È 1 È 5 ti 2 ”
Compsognathus longipes 0 0 0 0 0 0 0 1 0 i; 1 1 0 D: 0
Sinocalliopteryx gigas 0 0 0 0 0 0 0 Le (0) 1 di a 0 1 pe
Huaxiagnathus orientalis w i 0 0 0 0 0 0 0 0 5 IL 0 1 4
Sinosauropteryx prima 1 0 0 0 0 0 0 0 0 0 dl di 0 4 0
Scipionyx samniticus 0 0 CIO O 1 0 NL 0 Il gl di 0) ? ?
Deinocheirus mirificus 2 ? E E 6 ? ? ? ? Gi Vi 2 E; ti L
Harpymimus okladnikovi 0 Di Di z Z 0 JE iL (0) 0 1 Il (0) al vi
Pelecanimimus polyodon e È dl È il 0 © 1 0 ? 1 si 0 0 2
Shenzhousaurus orientalis p 0 1 0 2 0 e 1 0 0 & 9? 0 ? 1
Archaeornithomimus asiaticus 0 0 ? Gi ? f; È; ? fa 2 n f; ? ? ?
Garudimimus brevipes ? E ? (0) ? ? îl dl 0 0 1 1 0 N 2
Anserimimus planinychus Pi 2 ? d 9 È È E $ ? ti ? E È ?
Ornithomimus edmontonicus 0 ? f; 0 fi; ? si sh 0 0 il al 0 dì 2
Struthiomimus altus 0 ? ? 0 fi fi al il 0 0 al al 0 gl 2
Gallimimus bullatus 0 0 C; 0 fi fo il 1 0 0 il 1 0 il 2
Falcarius utahensis 1 al 1 il 1 i ? È 2 0 fi 2 n ? td;
Beipiaosaurus inexpectus ti È; 2 È 1 E ? ? p 0 ? ti Pi ? ?
Alxasaurus elesitaiensis 5 È; 1 il 1 Il È 6 ? Hi ? ? ? d Ts
Nothronychus mckinleyi al L di È dl ? Vi; ? È; ? D È; ? È ;
Erliansaurus bellamanus il ? ? ? IE È; E d ? è ? ? È ? ?
Nanshiungosaurus brevispinus 5 ? g 2 E; ? È 6; op 0 t; ® fa d è
Neimongosaurus yangi 1) ? 1 2 il ? ? 2 ? fi ? ? ? ? ?
Segnosaurus galbiensis 1 0 i 2 IL Nt IL R ? 0 ? ? ? fi ?
Erlikosaurus andrewsi 1 7; IL È L il ? 0 0 ? 1 0 10) 1 ?
Therizinosaurus cheloniformis 1 ? ? se F: f; & ? È 2 ? f: ? ? ?
Alvarezsaurus calvoi e ? ? 0 ? 7. g ? ? 0 ? ? ? ? A
Patagonykus puertai sì 0 ? ? ? ? ? p ? ? ? ? a ? ?
Mononykus olecranus 1 È ? 0 0 0 ? ? fi ? ti fi 0 di 2
Shuvuuia deserti Tì ? n 2 ? 0 ? 1 o) 0 ni 1 0 ? ?
Incisivosaurus gauthieri ? È 0 0 1 (0) P 0 0 ? 1 0 È il 0
Protarchaeopteryx robusta D 2 1 È. 1 0 1 ? È 0 ? 0 fa ? %)

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa v
Avimimus portentosus
Caudipteryx
Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis
Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi

Atrociraptor marshalli

Epidendrosaurus ningchengensis

Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

Characters

0
(0)
0

OOo ao e

en ss Ke) ie ke ie)

"Vv

*V

(ES ES Ot SS) st) let ela

*v *v

*V etto (Pe c&O' Si CORIO

*v

N_N

USI

ei die) demo ie fem def e ti,

*V

0
0

Ta

2268221002288 2298230823182328233823482358236023,/823882398240

249

250

Taxa / Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

241 242 243 244 245 246 247 248 249 250 251 252 253 254 255

0 0 0 0 0 0
0 0 0 0 0 0
al 0 0 0 0 0
1 di ? ? 0 2
Il 1 0 0 0 0
i D 0 0 0 0
0 0 fa n 0 ?
F D 5 ? ? DI
? 0 1 ? ? 1
l 1 0 0 0 0
2 2 ? PR DI DI
? 2 5 ? 2 ?
7: ? ? p) ? ?
? ? ? ? ? p
? ? D ? ? D
? 0 Ni 0 0 1
P ? ? 7) 7) ?
? 0 il 0 0 1
0 0 0 0 0 1
2? 0 Il 0 0) 1
0 0 L 0 0 1
È; 0 ci Il 0 1
2? È 2 2 P 7)
0 0 1 0 0 3
0 ) di 0 0 1
0 0 ? 0 0 Ù;
2? 2? ? ? 2 2
0 0 di 0 0 i
? ? z ? ? D
0 0 1 0 0 sl
0 0 l 0 0 1
0 0 il 0 0 1
? ? 3 ? p »
? 2? 2 ? 2 e)
2 2 È ) 2 2?
? 5 È ? ? ?
2 9 ? ? ? D
? R È ? È 2?
z 2? 2 7) ? ?
2 2) ? ? ? >
0 0 sl 0 0 1
2 2? ? ? ? D
e. ? B ? T; 2?
? ? ? ? ? ?
2 ? 7) n ? ?
ONTO Ji (0) 0 ii
: 0 al 0 0 il
2? 2? TI 2 0 9

0
0
0
0
0
0
0

|

0
0

e] xeol eo)

0
0
0
te
0
0
?
?

0
0

*v

(ele 103)

0
0
0
&
0
0
2

?

0
0

0
0
0

Gi

(©) =

0
0
()

2

(ee een)

0
0
n
t
0
0
f;

È

0
0

0

0
0
0
0
0

*v

Sd DO

DD a a 3 O 5

1-23> DR <® MIR

(N°)

0
|,
Ci

È

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

241 242 243 244 245 246 247 248 249 250 251 252 253 254 255

(e \atie = Mi;

*v

iena e)

(e> fn oe)

DAS

Si CD

»\

[a deh teft e’

*v

0
1

?

1
1

?

1
Vi

È

ti

ci soy «ef en (emo) xe iii ©)

*v

p

0
O)

È

1

251

DI

Taxa / Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polvyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

256 257 258 259 260 261 262 263 264 265 266 267 268 269 270

al 0 0 0 0 0)
1 0 0 0 0 0)
1 1 Di 0 n 0
1 dl di t; 0) 0
il il 0 0 0) 0
1 Ù 0 0 0 0
2 n 2 2 Di e
c; 6 5 fo ? 5
0 0 ? 0 0 0)
0 (0) 1 0 ? 0
d k 3 È fi P
7) D 2 z ? ?
p ? 2 2? 2 ?
? p. ? È a ?
z ? = 2 D ?
2 0 = 0 2? È
2 ? > ? È ?
Oda LO) TA 0 0 0
0 1 fi È) 2 0)
E; al = 0 L: 0
? 0 i 0 fs 0)
È 0 53 0 È (0)
? È ? 2 2? ?
0 È È È 0 0
0 0 = 0 1 0
e) 2 2 2 2 0
? 9 a 2 2? e)
0 ? 9 2 e) e)
n 7 7) 2 2 ?
? 2 E; 2 2 2
0 ? 2 2 © 2?
0 Fi 2 2 È ?
È D È 0) 0 0
? 2 A ? 2 2?
E È 0 t; 0 1
? ? 0) ? 2? È
2 ? ) D 2? P)
? 2? 2 n ? ;
2 ? ? 2 o) 1
? ? ? ? 0 1
0 0 0 0 0 L
È ? 1a ? ? ?
? 2 D 2 2? D
2 p ? e) ? P)
? 2 ? o) 2 2
0 f: n 0 ? 0
0 E = 0 0 di
5 5 ® R 0 sl

0
(0)
0

ti
0
0

&

0
0
0
0
0
0
|

1
Al

?

0

0

0

DS)

0

fi

0

?

0
ti
È

4

0
0

0

(0 E ce i DA o.

0

& ‘(© 5 PP _ © (o.

"Vv

9 a

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa /
Avimimus portentosus
Caudipteryx
Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis
Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi

Atrociraptor marshalli

Epidendrosaurus ningchengensis

Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

Characters

(> Jia © NRE |

v

“vv

*)

*)

0
0
0

0
0

?

(00 —_ I © ST ©

il
i
al

2
il
1

&
i
1

(A

È

0

sa

G
È;
?
È

?

?
IL
?
È

È

2560257258259 72607261 2627263264265 266 267 2682697270

2
0
L:

È

53

254

Taxa / Characters
Allosaurus fragilis
Sinraptor

Dilong paradoxus

Eotyrannus lengi
Tyrannosaurus rex
Gorgosaurus libratus
Tanycolagreus topwilsoni
Coelurus fragilis
Ornitholestes hermanni
Guanlong wucaii
Aniksosaurus darwini
Nedcolbertia justinhoffmani
Nqwebasaurus thwazi
Santanaraptor placidus
Orkoraptor burkei
Juravenator starki
Mirischia asymmetrica
Compsognathus longipes
Sinocalliopteryx gigas
Huaxiagnathus orientalis
Sinosauropteryx prima
Scipionyx samniticus
Deinocheirus mirificus
Harpymimus okladnikovi
Pelecanimimus polyodon
Shenzhousaurus orientalis
Archaeornithomimus asiaticus
Garudimimus brevipes
Anserimimus planinychus
Ornithomimus edmontonicus
Struthiomimus altus
Gallimimus bullatus
Falcarius utahensis
Beipiaosaurus inexpectus
Alxasaurus elesitaiensis
Nothronychus mckinleyi
Erliansaurus bellamanus
Nanshiungosaurus brevispinus
Neimongosaurus yangi
Segnosaurus galbiensis
Erlikosaurus andrewsi
Therizinosaurus cheloniformis
Alvarezsaurus calvoi
Patagonykus puertai
Mononykus olecranus
Shuvuuia deserti
Incisivosaurus gauthieri

Protarchaeopteryx robusta

CRISTIANO DAL SASSO & SIMONE MAGANUCO

271 272 273 274 275 276 277 278 279 280 281 282 283 284 285

Dior ng ee EVO
n 0 ? pi s, i,
Vi 9 2000 ia;
pe ? x) e) Al i
Guetta 0a
i ul: “0° celo
gi 00% Lote da a
cage, ST Sl a
Ca ae e N
Sa Cab
a an: le i
n a Ni p; n o”
I allo la IR RO
È ? ? ? ? ?
? ? ? ? 5 ?
O°. ce e TONO
D ? 7) ? È p
Gi oe, posò dia: Radio
dt La eine
CO IR CN a
gi apo
OP gi LT ara
) 0 2? DI dk 0
n 0 2 n 1 0
°°, Tg TG
Go) ? ? 2) mo n
Pep, ea
2 pe 2) 2? ‘DI >]
dl a Ra
È ed Co gine
È SOM
(2900 e o
Q po) ti n ci ni
iP) 1a 0 p. 2 (P)
2 i n
n e
=. Cola iena
2 P) DI 2? n? p)
ne E aaa
è è COPIO:
? ) ? È 0 ?
8 QAR
) E; P) ? n e.
PÒ p 9 7 Di ?
è 0. Ar 0a
È Lio ila, osa
vr) n? SD n ? n

0

OOO RDSTTSO Oa.

*v

0

T

0

ati Se di es)

0

0

DOS)

*v

WE UDIRE IO RS LATO

Pe ITO!LD9IEEa nto

DI OD ae Mea

0

‘> MN L= MIN 93 o O o DR o)

*v

; = MID

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

271 272 273 274 275 276 277 278 279 280 281 282 283 284 285

?

n

9

0

*V

*v

IO TATRO RO TO IDO? Pu

*v

sa
0
0

È

p_K LE HO _ HW

*v

NC N -_ NUNC DON CN iv

*"V

1
1
al

p_ht RP E HI_HEE "| HH «PF

NS)

il
al
Il

*v

ACRI T Rao

*v

?

ol
il

*v

DOES CNSITIOIT TOI II

*v

?
0
0

5

e)

i E La E (0e ei = N ©»; o

*v

1
sl

G

p_pno HE

O H H HW HE HW

*v

p_HE HE HE H KH Ho O v

*v

SOTTO.

*V

“Vv

RSS TORO CP LOLeO

*v

"v

p_H_ po HE HO EP HW H_H

*v

*v

Dr lo. I i ‘©

255

256 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Taxa / Characters 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300
Allosaurus fragilis OT 00 0; ee REGATA LO 00) 0
Sinraptor 0 0 ? 0 ? È P ? ? ? È È 0 0 R
Dilong paradoxus o o deo de Le
Fotyrannus lengi o ea i e e
Tyrannosaurus rex O. 0, È O 06, e 0 ia ee
Gorgosaurus libratus 1 0 Di 0) 0 0 & 0 È 0 0 0 0 0 0
Tanycolagreus topwilsoni 0 0 i 0 0 0 si 0 0 0 1 0 0 0 0
Coelurus fragilis È ? 3 0 0 0 1 0 0 0 È 2 ? ? È
Ornitholestes hermanni È È Si 0 e ? fs d 2 0 0 0 1 0 0
Guanlong wucaii 0 0 1 0 0 0 di 0 0 0 0 0 0 sl 0
Aniksosaurus darwini G È d È È ù 5) ti a ? 0 ? 0 ? ?
Nedcolbertia justinhoffmani È Gi t È È È 5 È È ? 6 ? ? ? R
Nqwebasaurus thwazi 0 0 2 0 0 î 0 0 0 0 0 sang x 0
Santanaraptor placidus È 6) 5 È 5 È ? a 2 ? ? È C f ?
Orkoraptor burkei È È di E io f; È 6 5 È fe È ? fi; é
Juravenator starki E 3 1 0 0 0 1 0 0 0 0 0 1 al 0
Mirischia asymmetrica È Gi È È È 6 È ? ? 2 ? ? ? de 8
Compsognathus longipes il 0 di 0 0) 0) al 0 Ki 0 0 (0) 1 1 0
Sinocalliopteryx gigas 1 È 0 0 0 0 1 0 |) 0 0 0 0 1 0
Huaxiagnathus orientalis 1 0 Il 0 0 0 Il 0 È 0 0 0 0 1 0
Sinosauropteryx prima 0 0 al 0 0 0 al 0 0 0 0 0 0 al 0
Scipionyx samniticus 0) 0 1 0 0 0 al 0 0) 0 0 0 Di 1 (0)
Deinocheirus mirificus 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0
Harpymimus okladnikovi 0) 0) Li 0 0 0 0 0 0 0 0 0 0 1 0
Pelecanimimus polyodon sl i Ji 0 0) 0 0 il 0 0 0 0 1 2 0
Shenzhousaurus orientalis (0) 2 al 0 ? 0 0 il (0) ? 0 0 il 0 ?
Archaeornithomimus asiaticus 0 0 1 0 0 0 È 1 2 ? 0 0 ù 2 0
Garudimimus brevipes Si 5 S È ? È ? d fi ? ? ? ? ? 2
Anserimimus planinychus al Il 1 0 0 0 Il 0 0) 0 0) 0 2 z 0
Ornithomimus edmontonicus al Di 1 0 0 0 1 1 0 0 0 0) 2 2 0
Struthiomimus altus 0 di si 0 0 0 0 di 0 0 0 0 Il Al 0
Gallimimus bullatus 0 0 i 0 0 0 0 dl 0 0 0 0 1 ì 0
Falcarius utahensis 0 0 1 0 0 0 Li 0 0 0 0 0 0 0 L
Beipiaosaurus inexpectus 0 0 at 0 È; 0 1 0 s 0 a 1 0 0 d
Alxasaurus elesitaiensis 0 0 È 1 E; f 8 È E È il 1 0 0 1
Nothronychus mckinleyi ? 6; 8 n ? ? ? E D ? ? ? (0) 0 0
Erliansaurus bellamanus 0 0 0 il 0 i 0 0 0 0 1 1 0 0 (0)
Nanshiungosaurus brevispinus ? ? ? ? ? (4 ? ? 2 ? ? ? ? ? ?
Neimongosaurus yangi Ci; 7: È t; : ? ? fr t; È 1a Ci; 2 f V;
Segnosaurus galbiensis E; 2 ? E; f p: È % ? Fs Pe f: ? ? fi;
Erlikosaurus andrewsi ? ? D ? ? È ? ? 2 ? Di È È; ? 2
Therizinosaurus cheloniformis 0 0 0 1 0 1 ? 0) n È A LL fs i 0
Alvarezsaurus calvoi È; e ? fa ve ? ? Pe fo ° 0 ? 1 ? Vi;
Patagonykus puertai Li 2 2 0) ? 5; fi ? 3 0 0 ti al ? A
Mononykus olecranus al (0) 0 0 pi f; C 5 ft: 0 0 f; al D v
Shuvuuia deserti 1 ? 0 0 0 dI ? 0 Ù 0 0 0 il 1 0
Incisivosaurus gauthieri ca ? ? ? ? ? IC; 2 ? L A €; È 2 A

Protarchaeopteryx robusta Ts Fi; 1 (0) Li 0 di 0 0 0 il 0 0 0 0

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

286 287 288 289 290 291 292 293 294 295 296 297 298 299 300

*v

"

SISSI TOROTOo tO

È

0

Vs

0

*v

*v

for iof Leica feta

E

0

6;

0

DS)

*

O HE HE Hl ' _ HEH-HEEO O_O°OO

È

0)

"vo

e)

(Sie dr celti i ie)

G

al

2?

0

DS)

*v

GOT LoOo ro o to Ora

?

?

f* it (ITO' G S

G

0

Oa al ti. O OS

È

0

(ola = opa piena)

(©) DD *v

"Vv

Hp_H HÈ HE HE OO O O

?
0
0
3

1

HW HKE RP HP O FE Hi O cv

*v

?
di
E:
0
0

[Pie > = 00) SI Ra:

° (>) *)

SS)

SONS TOoOrTLo no Tt>o rOoTro

?
IL
0

La

(e Ri > a i IR A,

De) I up) "v

*v

(TT geo Rao) coi RR A

ts
0
?

A

257

258 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Taxa Vi Characters 3013027303830 48305830683078308 830983108311 83128313 8314831005
Allosaurus fragilis 0 0 0 0 0 0 0 0 0 0 0 Des 0 0
Sinraptor S P 0 DI 0 0 0 0) 0 0 0 0 0 0 1
Dilong paradoxus 0 L 2 0 0 0 a 0 1 0 ® 1 0 1 pi
Eotyrannus lengi ? ? f: E ? P E Gi È se ? ts ? 2 È
Tyrannosaurus rex Ki E 0 1 0 0 il 0 0 1 0 ci 0 1 1
Gorgosaurus libratus È 5 0 Ii 0 0 1 0 0 1 0 1 0 1 1
Tanycolagreus topwilsoni 0 al G ? a) E; f; 0 1 al 0 1 0 1 al
Coelurus fragilis % È e Te ? ? n 0 LE 1 TA 1 0 ni Si
Ornitholestes hermanni È Cs 0 al (0) 0 il 0 e (0) È 1 0 0 ?
Guanlong wucaii 0 1 0 0 0 (0) 1 0 1 1 ? al È al 1
Aniksosaurus darwini i ss 5 z 6 2 È È il 1 e Dl È dì ?
Nedcolbertia justinhoffmani & {e COZZE 0 ? 0 ? al 0 ch ? 0 ?
Nqwebasaurus thwazi 0 0 ? ? ? P fi È al Fe È fi % al ?
Santanaraptor placidus 2 È 6 6 il 0 al ? ? di ? ci E sl ?
Orkoraptor burkei 6 E 5 e E Gi ? E ? G 0 ? n ? ?
Juravenator starki 0 0 0 ©. è È ? È 1 ? ? 1 t, ? 1
Mirischia asymmetrica G n ? 1 0 0 0 0 6 1 % D i ? ?
Compsognathus longipes & Ci 0 il 0 0 1 0) al ? Ta 1 0 al 1
Sinocalliopteryx gigas 0 at 0 il 0 0 1 0 1 ? ? 1 0) dl 1
Huaxiagnathus orientalis 0 fi 0 1 0 0 1 0 Di È 2 1 0 il di
Sinosauropteryx prima 0 0 0 Al 0 0 1 0 1 0 ? 1 fa il "
Scipionyx samniticus 0 Il 0 0 0 0 il 0 ? ? 0 ? ? ? ?
Deinocheirus mirificus 0 al 9 a È ? ? ti ? ? ? ? ? si ù
Harpymimus okladnikovi 0 2 0) ? È ? $a 0 ? ? ? ? 0 1 È
Pelecanimimus polyodon 0) Z È fs 2 îi ? È 2 È È ? ? Le È
Shenzhousaurus orientalis 0 Z 0 0 ? 0 ? 0 ? 0 ? ? ? ? ?
Archaeornithomimus asiaticus 0 E 7 0 0 0) di 0 1 0 0 È si ol ?
Garudimimus brevipes i È 0 2 E fi; 4 0 il 0 1 iL 0 di L
Anserimimus planinychus 0 sh E ? G ? E È É E ? ti È: sl ?
Ornithomimus edmontonicus 0 Z 0 0 0 0 i 0 ti 0 z 1 1 il 9
Struthiomimus altus 0 2 0 0 0 0 1 0) 1 0 ? 1 1 1 l
Gallimimus bullatus 0 2 0 0 0 0 1 0 1 0 1 1 di 1 ?
Falcarius utahensis 0 0 0 2 dl ? Il 0 ul ? ? ? ? 0 2.
Beipiaosaurus inexpectus 0 È 1 É 5 & E È 1 1 È 0 a 3 R
Alxasaurus elesitaiensis È b; 1 ? 0 0 I ? ? e; 0 0 ? ? fs
Nothronychus mckinleyi B E ù P 0 0 al È 0 ? Ci 4 ? Ù ?
Erliansaurus bellamanus 0 al È E 5 7 È È 0) 1 È È È È; Ls
Nanshiungosaurus brevispinus 6 8 1 0 0 1 1 ? Gi ? n ? v P È
Neimongosaurus yangi fs d 2 ? ? te fi ? 0 1 0 0 0 0 È;
Segnosaurus galbiensis È ? 1 0 0 0 I 0 0 Li Ca 0 È 0 l
Erlikosaurus andrewsi ? ? 2 ? È ? ? A È 5 fi 0 ? 0 ?
Therizinosaurus cheloniformis Ss È È È E 5 È E 6 © È n È e ti
Alvarezsaurus calvoi D ? 0 ? © f- ? È ? ? ? fs ? 1 ?
Patagonykus puertai È ? % ? È tO ? 5 fi 1 n ? fi I È
Mononykus olecranus ? ? Li » ? È ? E il 1 1 È ; il a
Shuvuuia deserti 2 ? 2? 0 ? s 5 Li il 2 E 1 I di ?
Incisivosaurus gauthieri È ia ? ? ? D 4 ? ? 7; C, f È ta E;

Protarchaeopteryx robusta 0 0 0 0 0 0 1 0 Li 1 1: 1 d I ?

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

301 302 303 304 305 306 307 308.309 310 311 312 313 314 315

Ci

1

ta

*v

*v

D. o EP eo ao o_o

L:
5

?

&

(«> NIE >|

0
0

È

0
0
0

1
1
sl

P_ FE PH RE HH HE

*v

1
l
I}

?
è
È

P

si
IL

?

p_pF LP HP _H _ HR HW _H

*v

0
0

&

1
il

?

"v

[NO]

po_pPo_ Ho _ HE _ HE KH H

*)

*v

p_pHh H_HR HE HH HW

*

®

1

©

PSI

260 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Taxa / Characters 316 317 318 319032073218 3223238324325 0326032:/0328083 298330
Allosaurus fragilis 0 0 0 0 0 0 0 0 0 0 0 (0) 0 0 0
Sinraptor 0 ? 0 0 0 0 0 0 0 f; p 0 ? 0 (0)
Dilong paradoxus 0 0 0 0 0 0 0 0 0 (0) 0 al 0 ? ?
Eotyrannus lengi 2 E t; Ù t; È B È E E ? ? 1 ? 1
Tyrannosaurus rex 0 ? 0 0 0 0 0 0 0 E ? ? 0 0 0
Gorgosaurus libratus 0 0 0 0 0 0 0 0 0 5 5 ? 0 0 0
Tanycolagreus topwilsoni 0 0 0 0 0) 0 0 0 0 0 0 al il ? ?
Coelurus fragilis 0 E È E È n ? ? 2 E fa 1 di ? ?
Ornitholestes hermanni 0 î È £: È f; ? ? 0 ? ? ? ? 0 0
Guanlong wucaii 0 È 6 2 2 Il (0) ? 0 (o) 0 () 0 ? 0
Aniksosaurus darwini 0 2 0 0 0 1 sE 0 2 2 ? ? ? p. 0
Nedcolbertia justinhoffmani 0 5 0 È 0 0 t; 0 0 G Le ? ? ? ?
Ngwebasaurus thwazi 0 0 0 0 0 0 ? 0 0 1 1 0 0 ? E)
Santanaraptor placidus 0 3 t: Gi ? E; ti ? 9 ? $ $ ? 0 0
Orkoraptor burkei fe È È G È n Z fo ? E È € ? ? D
Juravenator starki È 0 0 0 0 1 0 0 0 0 0 0 0 ? ?
Mirischia asymmetrica E z È È F; E È PR È d ? g: E; ? 0
Compsognathus longipes 0 0 0 0 0 0 0 0) 0 ? e 0 0) 0 0
Sinocalliopteryx gigas 0 0 0 0 0 0 È ? 0 0 0 0 0 0 0
Huaxiagnathus orientalis 0 0 0) 0 0 0 0 ? 0 0 0 al 0 0 0
Sinosauropteryx prima 0 0 0 0 0 0 0 0 0 0 ? 1 0 0 0
Scipionyx samniticus E 5 È E E E i i ? 0 0 0 0 0 0
Deinocheirus mirificus È È; È É È È; È È E af 1 0 0 ? 2
Harpymimus okladnikovi 0 ? È ; 0 0 0 0 0 0 di 0 0 0 0
Pelecanimimus polyodon È 6 È a E fo 5; È 2 di al 0) 0 È ?
Shenzhousaurus orientalis È Gi E È È 5 Gi È £ 1 0 0 ? 0) R
Archaeornithomimus asiaticus 0 5 E È A d E: È È ? ? 0 0 0 0
Garudimimus brevipes 0 0 di 0 È ? ? 0 (0 n fa R ? ? ?
Anserimimus planinychus 0 ? ? fi; ? È ? 7 E ? il 0) 0) ? ?
Ornithomimus edmontonicus 0 È s 5 1 0 0 0 è) il l 0 0 0 0
Struthiomimus altus 0 È È E il 0 0 0 0) 1 LL 0 0 0 0
Gallimimus bullatus 0 G ti ? al (0) 0 (0) 0 i Il 0 0 0 (0)
Falcarius utahensis 0 ? 0 ? 0 0 Fi 0 0 0 0) il O) ? ?
Beipiaosaurus inexpectus (- 0 È; 1 t) t: È È, 5 0 È 0 0 F, C.
Alxasaurus elesitaiensis E È, E È 0 0 0 0 0 E È 0) s 1 6
Nothronychus mckinleyi ? E 6 C È ? È ti al D; ? PR G ch ni
Erliansaurus bellamanus ? 2? Ci Ri E di ? 6 ? 0 1 0 0 ? n
Nanshiungosaurus brevispinus ? ? ? ti; A ? ? ? 2 2 ? ? ? ? 0
Neimongosaurus yangi 0 0 g G; 0 0 0 0 L) È: C: È di ? £
Segnosaurus galbiensis 0 0 1 pi ? 7) tr: 0 1 & 2 È e ? ?
Erlikosaurus andrewsi 0 0 dl Il 0 0 0 0 1 ? ? E; ? ? Si
Therizinosaurus cheloniformis È È È Ss € 5 È Bi È È È È È 6; fs
Alvarezsaurus calvoi 5 E; fs f; 0 0 0 o 0 ? È ? ? ? Y.
Patagonykus puertai % 2? dI 0) Gi Vi ? ri ? ? ? E ? DI ?
Mononykus olecranus È 0 0 0 0 0 0 0 0 ti È f: ? (0) 4
Shuvuuia deserti 00 0 ? 0 0 1 0 0 1 1 0 0 0 2
Incisivosaurus gauthieri 2 2 & 2 ? ? È C a ? ? . ? E; E:

Protarchaeopteryx robusta 0 0 0 (0) 0 0 n ? 0 0 0 0 0 È È

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

1683178318 831983208321583220832383248325 8326032132373 298330

E Pòì rei ela Ta

"Vv

"

e ie I 0 co lic DR > I «DEI <> A,

"Vv

*v

H_ Po_pP HR _HKF HO _H _ HE_Òm

*V

0
O)

G

DI LD

birillo tlote

*V

*

Rie eri LOU OD

0
0

i;

[© i ©)

> RO > MRI I)

"vv

0
0

È

(esi ie

"Vv

(© (I)

s

0
0

?

0
0

?

SD

‘Vv

> Si. i fa COS SI 13

*v

*

*v

PES To T>TTOITIOo ii >

si
È

e

0
0

2?

‘vv

> SS Si © ix» CNAI( <°, (e e)

*v

[@)

*v

261

262 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Taxa / Characters 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345
Allosaurus fragilis 0 o 0 0 0 0 0 0 0 0 0 0 0 0 0
Sinraptor 0 0 0 0 0 ? 0 0 ? 5 0 0 0 ? ?
Dilong paradoxus 5) 0 2 0 l S 0 0 ? ? 0 2 0 5; 0
Eotyrannus lengi > £ ® € 1 ® @ 00000, soa
Tyrannosaurus rex 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0
Gorgosaurus libratus 1 0 0 0 l 0 0 0 0 0 0 0 0 1 0
Tanycolagreus topwilsoni il È 0 È 1 È È 0 0 0 0 0) 5 0 0
Coelurus fragilis I È 0 6 iL É ? ? 0 0 P ? 0 0 ?
Ornitholestes hermanni 0 0 (0) 0 il 0 0 0 0 0 È È 0 : a
Guanlong wucaii 9? 0 0 0 ah ? 0 0) e ? 0 0 0 0 0
Aniksosaurus darwini E 0 0 È 1 2 0 0 6) 0 0 0 ? 2 ?
Nedcolbertia justinhoffmani 0 È È D il 0) ? 0 È 0 ? 0 ? ? ?
Nqwebasaurus thwazi E fa È G LL ? ( 0 0 0 P 0 ? 0 1
Santanaraptor placidus È È È 0 z È È E È P È ? È Gi È
Orkoraptor burkei È È (i È fa È 5 & 4 i ? 2 di ? e
Juravenator starki B 0 0 È È 0 0 0 0 0 0 0 0 0 0)
Mirischia asymmetrica 0 ? É 0 5 ? ? ? e ? P ? ? ? ?
Compsognathus longipes 6 0 0 0 il 0 0 0 0 0 0 0 0 0 0
Sinocalliopteryx gigas È 0 E 0 sl 0 0 0 0 0 0 È 0) 0 0
Huaxiagnathus orientalis E 0) 0 0 al 0 0 0 0 0 0 0 0 0 0
Sinosauropteryx prima 0 E 6 0 I 0 0 0 0 0 0 0 0 0 al
Scipionyx samniticus 0 0 ? 0 g ? 0 È 0 0 n 5 0 0 0)
Deinocheirus mirificus È d È È È & È, È 0 0 ? P 2? il 1
Harpymimus okladnikovi ? 0 0 E 1 Pe 0 0 0 0 ? 0 0 0 0
Pelecanimimus polyodon È 0 & 0 È È % 6 ? È ? ? 0 a: A
Shenzhousaurus orientalis E 0 ? 0 È 0 0 0; È R ? P 0 ; fo
Archaeornithomimus asiaticus È 0 0 0 1 0 0 È 0 0 G & e 1 ni
Garudimimus brevipes 0 0 0 È 1 0 0 0 ? Gi 0 0 0 ? ?
Anserimimus planinychus 5 E E 5 2 to P 5 0 0 È ? ? 1 di
Ornithomimus edmontonicus fe 0 0 (0) Z 0 0 0 0 0 ? 0) 0 ol 1
Struthiomimus altus È 0 0 0) 2 0 0 0 0 0 È 0 0 1 N
Gallimimus bullatus 0 0 0 0 2 0 0 0 0 0 ? 0 0 al il
Falcarius utahensis 5 0 0 0 0 0 0 0 0 0 0 0 0 ? 0
Beipiaosaurus inexpectus b E È È t; 5 1 ds È di 0 6 0 6 0
Alxasaurus elesitaiensis É 0 0 0 2 0 al 0 0 0 1 ? 0 i: P;
Nothronychus mckinleyi ? ? 9 il È ? ti ? 0 0 e. o ? ? ©
Erliansaurus bellamanus E fi fi ? ? ti Il ? 0 0 E È ? 0 0
Nanshiungosaurus brevispinus ? 6 6 0 È di 1 6 È & 1 È È E È
Neimongosaurus yangi C 0 : ® 0 0 al sl IL 0 sl 0 0 ? E.
Segnosaurus galbiensis E 0 0 dl ? n Il Il 2 0 1 ta 0 E, A
Erlikosaurus andrewsi È È ? ? 0 P ? iL il 1 1 0 0 f: V;
Therizinosaurus cheloniformis È ? a E Ls ? 2 ? 1 1 n ? 2? 0 7;
Alvarezsaurus calvoi fi 0 2 ? 1 0 0 0 ? Di K 0 Ò È ?
Patagonykus puertai a ? 2 ? ? fa 2 6 ? ji ? fe Le d 2
Mononykus olecranus fs 0 O) fi: ? È; 0 0 1 1 0 0 È 1 ?
Shuvuuia deserti 00 0 0 2 R 0 0 1 1 0 0 0 xi 2
Incisivosaurus gauthieri È ò ? Ci ? ? ? of ? ? ? È; 0 È: ad
Protarchaeopteryx robusta 0 0 0 0 1 0 0 ? 0 (0) 0) 0 0 |) 0

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

331 332 333 334 335 336 337 338 339 340 341 342 343 344 345

0
0

F;

0
0

C:

2
il

?

*V

O)

Nu. A © (nia ei

*v

N

"VO

[Ma

P_H P o_E PR _N N E

*V

FP)

0

P

0
0
0

0

tei i:

")

*v

(elio SX} Meme e ARE,

v

(0)
0
0

*V

SOTTO TROoOro o Lo u9

*v

0
0
0

*v

*)

STORTO o OD

“Vv

È

ot Mo) ao e)

[o Mesi ef eri eco co ee e

"Vv

*v

OM To Oo OLO 19

"

0
0

È

De

*v

*v

OMO TSO To Io Di.

*v

dd Po _P_ Po_H H_ O

STRO TOornMi e

| = ATO «=> IT > TE co PINO <> RI > O > RI © ET

*v

e)

OTO. to: =. 0° et I

"Vv

*v

*v

SY To r5 5 ST: (DD

263

264 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Taxa 7 Characters 34602347348 0349 0350035108352 0353835493550356835 0358089598560
Allosaurus fragilis 0 0 0 0 0 0 0 0 00 0 0 0 00
Sinraptor 0 0 al n 0 0 0 0 0 0 0 0 0 0 (6)
Dilong paradoxus 0 0) 0 sì 0 0 2 0 2 ? ? 0 0 0 0
Eotyrannus lengi 0 ? 0 È ? ti 0 0 Li p G ? ? ? ?
Tyrannosaurus rex 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0
Gorgosaurus libratus 0 0 0 1 0) 0 0 0 0 0 0 0 1 0 0
Tanycolagreus topwilsoni È È 0 Si 0 0 0 ? ? ? 7. 6 0 ? t
Coelurus fragilis fo 5 È È È 0 0 ? ? ? È 0 () 0 ?
Ornitholestes hermanni 1 1 0 ni È 3 0 0 0 0 0 0 (0) 0 0
Guanlong wucaii 0 0 0 ? È 0 e di p 2 0 0 0 ? 0
Aniksosaurus darwini È ? i È po [ 53 Di ° n <; ? 0 0 ?
Nedcolbertia justinhoffmani È; £ 1 È ? E gl ? 2 ? ? È 0) 0 ?
Nqwebasaurus thwazi È A 1° È ? Le 2 0 ? fi È OO ? ?
Santanaraptor placidus È È z o È ? 6 (i ? fa fi f: 0 E, ?
Orkoraptor burkei E È E & 5 Es ? 6 ? ? 2? i ? si 2
Juravenator starki 0 1 0 1 G 0 s 0 0 0 0 (0) ? ? -
Mirischia asymmetrica A È È È Gi È È Z fi & È fi; a ® ?
Compsognathus longipes 1 1 È 0 wi 0 0 0 0 ? 0) 0 0 0 =
Sinocalliopteryx gigas 0 0 0 di fs 0 ? 0 0 ? 0 0 ? ? -
Huaxiagnathus orientalis pi È 0 0 0 0 ti 0 ? Ri ? 0 0 0 -
Sinosauropteryx prima 1 I 0 0) 0 0 Gi 0 È ? ? 0 0 0 =
Scipionyx samniticus L l 0 fo 0 0 ? 0 0 0 0 0 ? Li >
Deinocheirus mirificus A È 1 È 0 E È È ? E; fe ? PD È ?
Harpymimus okladnikovi C R al L 0 ? 0 0 ? 0 ? 0 0 0 0
Pelecanimimus polyodon il di 1 ? 0 È ? 0 R 0 0 0 È ? 0
Shenzhousaurus orientalis fe li I E 0 Ti P 0 2 0 ? ? ? 0 0
Archaeornithomimus asiaticus È fi nl È 0 0 Al & È î È 0 ul 0 ?
Garudimimus brevipes È 1 È È È È 0 0 0 0) 0 0) 0 0) 0
Anserimimus planinychus i È 1 È 0 È È ù E ri ? ? i E; ?
Ornithomimus edmontonicus fe I 1 1 0 0 È ? 0 0 0 0 1 0 0
Struthiomimus altus ? 1 al al 0 0 Il 0 2 0 0 0 al 0 0
Gallimimus bullatus ? JI di gl 0 0 pi 0 0) 0 0 (0) si 0 0
Falcarius utahensis 9 ? 0 E 0 È; È R ? ? ? ? 0 0 ?
Beipiaosaurus inexpectus È E 0 2 0 d; ? Si ? 9 di tf; 0 0 È
Alxasaurus elesitaiensis È 6 0 0 0 5 ? A ? Ss ? fi ? 0 È
Nothronychus mckinleyi f; di 0 B E È G £ ? ? t; 0 E ? f;
Erliansaurus bellamanus È a 0 t: il w A ? S È 0 È; L ® ?
Nanshiungosaurus brevispinus fi È ci 5 È È G; ? É ? ? 0 ? ? (;
Neimongosaurus yangi ? ? ? 0 z 0 il ? A ? ? 0 0 0 d;
Segnosaurus galbiensis fi; ? C; ? G 0 1 È ? D P 2 0 vi: 0
Erlikosaurus andrewsi V; 1 È È € £ È 0 l 0 0 E A La 0
Therizinosaurus cheloniformis È o 0 È al 0 È È È È ti s È: E ci
Alvarezsaurus calvoi ? ? ql 1 ? Ù di ? da 6 ° 0 0 0) 2
Patagonykus puertai D 2 a ? E; ? (; ? f; Fs % fi; 0 0 E;
Mononykus olecranus ? È al È 0 0 ? 2 R ? ? 0 1 0 i;
Shuvuuia deserti E. 1 1 0 Di È 0 ? 0 ? 0 i 0 0
Incisivosaurus gauthieri i 1 2 n 2 €; ? 0 1 il 0 ? ? ? 0
Protarchaeopteryx robusta 1 1 0 1 0 fi 1) ? Te ? ? 7) 2 ? 0

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Taxa / Characters
Avimimus portentosus
Caudipteryx

Microvenator celer
Elmisaurus rarus
Chirostenotes pergracilis
Caenagnathus collinsi
Oviraptor philoceratops
Rinchenia mongoliensis
Citipati osmolskae

Zamyn Khondt oviraptorine
Ingenia yanshini
Conchoraptor gracilis

Khaan mckennai

Heyuannia huangi
Sinovenator changii

Mei long

Byronosaurus jaffei
Sinornithoides youngi

IGM 100/44

Troodon formosus
Saurornithoides mongoliensis
Zanabazar junior

Unenlagia

Buitreraptor gonzalezorum
Rahonavis ostromi
Bambiraptor feinbergi
Sinornithosaurus millenii
Microraptor zhaoianus
NGMC91

IGM 100 1015

Adasaurus mongoliensis
Velociraptor mongoliensis
Saurornitholestes langstoni
Deinonychus antirrhopus
Achillobator giganticus
Dromaeosaurus albertensis
Utahraptor ostrommaysi
Atrociraptor marshalli
Epidendrosaurus ningchengensis
Archaeopteryx lithographica
Wellnhoferia grandis
Jeholornis prima

Sapeornis chaoyangensis
Confuciusornis sanctus
Protopteryx fengningensis
Yanornis martini

Hagryphus giganteus

346 347 348 349 350 351 352 353 354 355 356 357 358 359 360

[eo I MO)

[o Moi eric fe del sie

*V

(a Meo)

Asolo ro e i OTO

DS)

O Too Orio oo DD LO

1
P
d

II

0
(0)

P

*v

"Vv

oo. loreto tolo ki OMO

USI

DI et 1)

SIIOMORO RIOT ROSS

(©) («D] *v *VJ (©) DO)

*v

0
0

5

OOO OTO LORO Ono

si gd

"Vv

1
0
0

il
0

?

CN

e

0

?

0
Al

265

266 CRISTIANO DAL SASSO & SIMONE MAGANUCO

APPENDIX 7
List of the sampling points for SEM analysis (Fig. 186)

1. Microsample of somatic musculature from the base
of the neck of Scipionyx, between the centrum of D2 and
the shaft of Dr1. It is likely a cortical (peripheral) slice, as
other layers underlay it.

2. Bone fragmentof Scipionyx, coming from the dorsal
wall of the shaft of Dr3, just cranial to the cranial margin
of the right scapular blade.

3. Dome-shapedelement(whose elliptical base has been
left in place), belonging to the nodule of calcite grains that
form a cluster in the ventral portion of the thoracic region
of Scipionyx.

4. Red macula in the thorax of Scipionyx: encrustations
on the shaft of the left radius, and on the matrix bordering
the cranial margin of the diaphysis of the right humerus,
provided 6 microsamples.

5. Cranialtract of the descending loop of the duodenum
of Scipionyx: microsample from the dorsocaudal section
of the tube, at the level of the right elbow.

6. Fold of the mucosa from the internal curve (cra-
nial margin) of the U-turn linking the descending and
the ascending loop of the duodenum of Scipionyx. In a
microsample from this point, even fragments of mes-
enteric and pancreatic tissue might be potentially in-
cluded.

7. Drop-shaped granule, positioned as the most cau-
doventral untouched element of the calcite cluster that in
the abdomen of Scipionyx parallels the ascending loop of
the duodenum, between D12 and S1. As for sampling n.
3, the element has been cut at the base.

8. Microsample from a right lateral section of the de-
scending loop of the rectum of Scipionyx, located between
the centrum of SS5 and the ischial feet. Given the irregular
preserved surface of the tube in this region, part of the
intestinal contents may be included in the sampling.

9. Fragment of one of the fish scales contained in the
faecal pellet that fills the end of the rectum of Scipionyx,
at the level of the caudal margin of the right ilium.

10. Microsample from the dorsal bundle of the left cau-
dofemoral muscle of Scipionyx. More precisely, it comes
from its dorsal margin, where it faces the cranial articular
surface of the centrum of Cal.

11. Sample of sediment from the granular bed emerging
below Scipionyx, at the level of the right tibia.

12. Sample of sediment from the granular bed emerging
below Scipionyx, at the level of the right femur.

13. Sample of sediment from the granular bed emerging
below Scipionyx, at the level of the left manus.

SUPPLEMENTARY INFORMATION

Character change list (ACCTRAN and DELTRAN) related to the data matrix published in Appendix 6
is available on the website
http://www.scienzenaturali.org/riviste/memorie/37sup.html

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 267

Fig. 186 - Map of the sampling points (numbered dots) selected for the SEM analyses published in this study.
Fig. 186 - Mappa dei punti di campionamento (pallini numerati) scelti per le analisi al SEM pubblicate in questo studio.

268 CRISTIANO DAL SASSO & SIMONE MAGANUCO

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY 269

Marco Auditore & Arianna Nicora, L'ultima cena (The last supper), Genova 2010. Matita su carta e Photoshop (Crayon on paper and Photoshop).

ANO DAL SASSO & SIMONE MAGANUCO

ti VIA, ULI,

SI
Sn Tele ii
PEMPMESI

n) 008

.

‘pia pes

Davide Bonadonna, Scipionyx samniticus a caccia di un Eichstaettisaurus (Scipionyx samniticus chasing Eichstaettisaurus), Segrate
2010. Tempera su carta e Photoshop (Tempera on paper and Photoshop).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY |

Paolo Cinquemani, Ciro nuota (Ciro swims), Trieste 2010. Acrilico su tela (Acrylic on canvas).

CRISTIANO DAL SASSO & SIMONE MAGANUCO

dboard).

ic on car

ino (Acryl

Ico su cartoncino

ili

LG
(9)
<
s
_—
(©)
N
(©)
[=|
ic
2
D
(9)
ta
sì
-_
HS;
SI
S
Sì
D
Pai
È
K
=
(S)
>
S
[SS
S
S
9
5
Q
dv
DAI
ka
È
©)
>
IZ]
o)
(o
so
2
Go)
sa

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

x

(9
N
»
LS
2

ZE
jik_4 (a

BA CITA,

Renzo Zanetti, A caccia nelle pozze (Hunting in pools), Monteveglio 2010. Acquarello e pastello su carta (Watercolor and crayon on
paper).

9, 74 CRISTIANO DAL SASSO & SIMONE MAGANUCO

__are OPS pira

TRO, AMINA

non x

caio
FEES AAA

PR 2.090

Loana Riboli, All'ombra della Zamites (Under the Zamites tree), Crema 2010. Tecnica mista su carta (Mixed technique on paper).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Fabio Fogliazza, Ricostruzione testa Scipionyx samniticus (Scipionyx samniticus head restoration), Piacenza 2010. C hina e acquarello
(Ink and watercolor).

A)

276 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Lukas Panzarin, // buongiorno si vede dal mattino (You can tell a good day from the morning), Torre di Mosto 2010. Acrilico su carta
(Acrylic on paper).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

UN

ei DICA

” ;
i AVRETE

«1

POSA GIA
TO

Tullio Perentin, Cinque piccoli ruffiani (Five little panders), Trieste 2010. Matita su lucido (Pencil on tracing paper).

CRISTIANO DAL S O & SIMONE MAGANU

Troco, Ciro e cicadeoidea (Ciro and cvcadoid), Venezia 2010. Olio su tela (Oil on canvas).

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

TABLE OF CONTENTS
IGO RONE een Pag. 6 Mypedlocali nn Pag.
Oro ASSETTINGEe een Pag. 7 Iypesnorizon'and'age nti. Pag.
Tic Lalla ERA Pace) Emendedidiagnosispo ea Pag.
Borca DacKSTOUN A nine Pago Remarks Rn Pag.
Geological framework...............-rericrecieneenee Pag 8810 OSTEOLOGICAL DESCRIPTION AND
Lithology, stratigraphy, and sedimentology ...... Pag. 12 GOMPARISONSEeee ee a ine Pag.
Stratigraphic position of Scipionyx ................. Paol (CNC AICaAlUres eater Pag.
Macroscopic aspect of the embedding SRUBIBANDIMANDIBLESESeEZA a n Pag.
ti e ore iiccogizonzicen Pag. 14 daddi RENE ESTA: Pag.
aaa x Pane oa naz UG ADEITUra DASI OSSEA Licio Pag.
MEA OVITONMENt ti ciiiinnn BacRis AGODA SA i Pag.
Becoonmodeli Mea Paolo SIOE Eat i e se
HALO basin model... PICIBIIG Maxilaraienestraee ne enna Pag
e models e DOOSINTO Ob e nin Pag
The fossil assemblage and palaeoenvironment . Pag. 17 Supratemporal fossa ................---rsrrrirn Pag.
WEeiipi ere BS Paogiti SUPIIempoa li oncsl drain i
HER, det RIS AT TRA Pag. 18 om Iter se:
IVEGEDralestso n nia Pag. 18 Mo sE
OSE a SSA RA Pag. 19 RIF ERI a] DE
e PIO Stelera plates enna Pag.
ee Pag. 20 Dermalshullicoofeee nn Pag.
AS Tnt southeni Italian Premaxllae eo ann Pag.
TOR See AI Pag. 21 MAU a Pag
Palaeobiogeographical remarks ................... Pace) LEE RAI Pag
MIRA Re Pag. 24 | RIN Se RAI AR Pag
WU o se REI Pag. 24 fiera een Pag.
Menini Pag. 24 Bislorpialeee ei Pag.
Optical MICrOScopy ............-srccrrcie eee rinenezese Pao 295 Iii Pag
Photographs and drawings ............................... Pas825 OM UA eee Pag
RIESSIIEMENIS o inn Pag. 25 Squamosileeee e ii Pag.
UNdiontanalysis i ie Paoae25 RIONGI eee ina Pag.
CLEMENS CIR Pag. 26 Poe i Pag.
SEM bay elia Pag. 26 Briatore Pag.
ATINICALIerMSte enne Pacs] Supraoconpitale = = Pag.
SWICMAlciis nn Pag. 021 Proccipiallee en Pag.
Anatomical abbreviations ................................ Pag o21 POR nn: Pag.
Insututional'abDreviations*t.......c-rrricnrrn Papae2) Basioccipital-Basisphenoid .......................... Pag.
PARTI-COSTEOLOGIYT. rn Pag. 34 Basisphenoid-Parasphenoid ......................... Pag.
Systematic Palaeontology .........................1 Pag. 34 Paierosphenoid iii Pag.
Type and only species ...............................-- Pag. 34 Orbitosphenoidi:iinnririnzerraneeee Pag.
Etymology sanno Pag. 34 Palatoquadrate complex .................................. Pag.
Holotype rina Pag. 34 VOM Ne, iena Pag

219

280 CRISTIANO DAL SASSO & SIMONE MAGANUCO

Palatineros eee Pag.
Pitelli Pag.
Epipieryeoido-=-= nea Pag.
Éctopterysod ee, Pag.
Quadrate an Pag.
Mandbleee == nre enne Pag.
Mandibularopenne sese nn Pag
Déntalyet == ee na Pag
Supradentary-Coronoldene atene Pag
Splenaleese eee o Pag
SUS Ae n Pag
ANQURIETEO Eee Pag
Preaticulare ee e e Pag
AINNCUORRERe O e en Pag
Dentitiol eZ se Pag
Premaallinyitech e ee na Pag.
Maxillarytechesese one een Pag.
Dentaryiceth ere ee Pag.
Héterodoniyi e ore Pag.
HY010d Appalti Uste e ee e Pag.
POSTCRANIALAXIAL SKELETON ............. Pag.
Vertebrae>: Rare ATO Pag
Cervicalivertechra fee e Pag
Dorsaliverebraeseeee e Pag
Sacral vertebra ee Pag
Caudal vertebra eee Pag
Hacematarches Ae e Pag
Ribs E Pag
Cervicalnbsete rare Pag
Dorsattrbsfe Pag
Sacraltbst SR Pag
Gastalla XS E e A Pag
APPENDICULARSKELEETONEze eee Pag
Pectoral ste e e Pag.
Scapila:/ Se Pag.
Coracold<.;... sn a Pag.
Furtila se CR e ea Pag.
Forli Pag.
HUMeEUs a RR Pag
Radius. ER SR Pag
Ulna ato e N Pag
Carpus RR A Pag
Metacarpust Ar Pag.

Manual phalansestae eee e Pag.

47
48
49
49
50
50
51
51
52
52
52
53
53
53
54
54
55
57
58
58
59
59
59
64
70
71
75
76
76
78
80
80
84
84
85
88
89
89
90
91
92
92
95
96

Pelvic.girdlest,.. ale Pag. 97
MU: Ra Pag. 97
PUDIS- 4. Pag. 101
Ischium Bit. I Pag. 103

Hindimbiferr aa Pag. 104
Femme nnn E Pag. 104
Tbla ne Pag. 105
Fibula = or Pag. 106

ONTOGENETIC ASSESSMENTER e Pag. 110
Introduction io e Pag. 110
Gut contents oe Pag. 110
Scarred DONE SUI Ces ts Pag. 110
Large skull re Pag. 111
Orbital foramina rounded and
proportionally very large and antorbital
regionisnort and'idecpe cn Pag. 111
Position of the maxillary fenestra ................ Pag. 112
Rostralramus'ofthe maulla ene Paglia
Unfused'interdentaliplaes eee Pag0li2
Nasals shorter than frontals .......................... Pag. 112
Position of the caudal margin of the
apertura NASIOSSCA RI e Pag. 112
Shape ofthe laica Re Pag. 113
Large prefrontals, with descending lateral
process well-exposed in lateral view ........... Pag. 113
Sublacrimal expansion of the jugal .............. Pag. 113
Frontoparietal fontanelle open ..................... Pag. 113
Elements of the braincase in loose contact .. Pag. 113
Vomers unfused to each other all along their
length. e Pag. 114
Craniocaudal length of the dentary .............. Pag. 114
Angular:moved dorsalyeee Pag. 114
Symmetrical tooth crown height in
counterlateral tooth rowsti een Pag. 114
Low:numberiof lateralteeth ee Pag. 114
Teethwithfewideniclest ee Pag. 115
Denticle:size and densiy e A Pag. 115
Incomplete ossification of the vertebral
columD i; TE Pag. 115
Cervical ribs not fused to the corresponding
Vertebraco ino i Pag. 117
Unfused sirdieelennb eee Pag. 117
Relative size of the girdle elements ............. Pag. 117
Degree of ossification of the ilium ............... Pag. 117
Non-ossified sternal plates ........................... Pag. 118
Limb proportions aree Pag. 118
Carpal count and ossification ....................... Pag. 118

SCIPIONYX SAMNITICUS FROM THE LOWER CRETACEOUS OF ITALY

Elongation ofthe manus..........................- Pag. 118
Development of the fourth trochanter .......... Pag. 118
Li Rie EE SIE Paosgkl9
BHNEDGENETICANALYSIS<......t Pag. 121
Modification of the data matrix ......................... Pag. 121
ae ie Pag. 121
(Ras Pag. 121
BI ninni Pag) 22
(Lui l e eee e iRet Pag. 123
BRGRGIAISIAPHONOMY?:...............-rrno Paos]2)
PART II - SOFT TISSUE ANATOMY........... Pag. 129
LIO RA Pag. 129
IMBEENASSOEBLTISSUES iti Pag. 129
BEEN WTTDIT NE DONEST nno Pag. 133
BEE INaINS o nnnnnnn Pap 195
MESI O MENSA nn Pag. 136
Beailoriicnlaricarilla gesti... cine Pag. 137
Appendicular articular cartilages ....................... Pag. 137
Muscles, connective tissue and other soft
cus MN Pri Eee Pag. 138
Tai at A O Ae Pag. 141
Oesophagus and stomach .................................. Pag. 142
Iborsallepaxial MUSCles...............- rin Pag. 143
Liver and other blood-rich organs ..................... Pag. 143
GORE SRI E Pag. 145
MIRINO nn Pag. 147
enna Pag. 148
eee nno Pag. 149
luci fe RReIoEO Pag. 150
Ber peli eee PapsglSi
Niesenteric DIOOd vessels .-....--:--..-.ccrearsraricerasco Pag. 153
elvicfand hindlimb muscléest...-...--.. rari Pap@el55
Emporschiofemofal Musclena.t nn Pag. 155
arnioiemoralimuscles tenne Pag. 156
Ilio-ischiocaudal muscle septa ..................... Pag. 157
Other fragments of pelvic and hindlimb
Misteri Pag. 161
Caudal hypaxial connective tissue/ligaments .... Pag. 161
Indeterminate ?connective tissue ....................... Pag. 162
EIERNALSOERTFISSOESE, in Pag. 163
Bomyiclays ne n Pag. 163
ENDOGENOUS BIOMOLECULAR
BOMPONENISE SE nonna Pag. 165
SORIMISSUESAPHONOMXYXe&=eeE se. Pag. 167
Diagenetic formations possibly related to soft
ASSISI O nani Pag. 169

281
PART III - FUNCTIONAL MORPHOLOGY
AND PALAEOBIOLOGY .............................. Pag. 171
SKELETAL RECONSTRUCTION AND
IN VIVO RESTORATIONS OF SCIPIONYX
SAMNIK COSE Pag. 171
Granialireconsuucuoneee o en Pag. 171
Axialiskeletoneteeee nn Pag. 174
Pectoralis;rdle'andiforelimbiz.. uo Pag. 174
Pelvic girdle and hindlimb ........................... Pagsli5
Body length and body mass ......................... Pag. 175
Meg uMenteeeee enna Pag. 175
GUT CONTENTS AND FEEDING
GHRONOLOGNEE ES e n o Pag. 176
QesophapealiCOnientstiiz....onnnrnne Pag. 176
Gastrle:coMmenSeeg nn Pag. 176
DEscipIioneeen Pag. 177
ilaxonomica lines Pao li9
Iptestimalicontienis eee ee o Pag. 179
Palaeobiological significance of the gut
CONIENIS'OLISCINIONID e Sette testi Pag. 181
REMARKS ON THE PHYSIOLOGY OF
SGIRIONIAEO O core Pag. 183
Digestive phySi0l0gy ..............rrriiiiiie Pag. 183
Respiratory phySiology ................. iii Pag. 185
liverfandidiaphra smesso Pag. 185
Diaphragmatic muscles ...........................0..- Pag. 186
IIDOSR i iena Pag. 186
Mocneaeeee iannn Pag. 188
findominalQgsacse nn Pag. 188
Osteological correlates: pneumatic bones .... Pag. 188
Osteological correlates: morphology of ribs
“ndnerteba ee et enna Pag. 189
Osteological correlates: gastralia ................. Pag. 189
Osteological correlates: pelvic girdle ........... Pag. 189
GONCEUDINGREMARKSErat e nn Pag. 190
AGENOWEEDGEMENIS EE Re Pag. 191
Specialthapks Rene Paol 92
REFERENGESTee e nn Pap 199
NOTLEADDEDUN:PROObBeza enna Pag. 205
APPENDI alpe na Pag. 207
APPENDEREeee e ina Pag. 212
APPENDIB Eee e Pag. 214
APPENDDG4E ee ino Pag. 215
APPENDIX5Eee cin Pag. 216
APPENDIXMOPe ee ie Pag. 218
APRENDIXMge an Pag. 266
SUPPLEMENTARY INFORMATION .............. Pag. 266

Il - MONTANARI L., 1969 - Aspetti * DA
d'Orta). pp. 23-92, 42 ie Pi parare del Lias di Gozzano (Lago
"i ‘olche n clin nà MI G. C. & DAL PIAZ G. V., 1970 - Ri-
suo substrato cristallino. pp. 93 Tappe geo (Prov. Torino) e sul
figg.. 4 tavv. a colori e 2 b.n. | con carta a colori al 1:40.000, 14

Volume XIX
1- CANTALUPPI G., 1970 - Le Hi/doceratidae del Lias medio delle regio-
nì mediterranee - Loro successione e modificazioni nel tempo. Riflessi
TEN e sistematici. pp. 5-46, 2 tabb. n.t.
Il - PINNA G. & LEVI-SETTI F., 1971 - 1 Dactylioceratidae della Provin-
cia Mediterranea (Cephalopoda Ammonoidea). pp. 47-136, 21 figg., 12

tavv.
Il - PELOSIO G., 1973 - Le ammoniti del Trias medio di Asklepieion (Ar-
SIRIA, Ping del «calcare a Prychifes» (Anisico sup.). pp.

Volume XX

1- CORNAGGIA CASTIGLIONI O., 1971 - La cultura di Remedello. Pro-
blematica ed ergologia di una facies dell’Eneolitico Padano. pp. 5-80,
2 figg., 20 tavv

Il - PETRUCCI F., 1972 - Il bacino del Torrente Cinghio (Prov. Parma).
Studio sulla stabilità deì versanti e conservazione del suolo. pp. 8/-/27,
37 figg., 6 carte tematiche.

MI - CERETTI E. & POLUZZI A., 1973 - Briozoi della biocalcarenite del
Fosso di S. Spirito (Chieti, Abruzzi). pp. 129-169, 18 figg., 2 tavv.

Volume XXI
I- PINNAG, 1974-1 crostacei della fauna triassica di Cene in Val Seriana
( 0). pp. 5-34, 16 figg. 16 tavv.
Il - POLUZZI A., 1975 - I Briozoi Cheilostomi del Pliocene della Val d’ Ar-
da (Piacenza, Italia). pp. 35-78, 6 figg., 5 tavv.
II - BRAMBILLA G., 1976 - I Molluschi pliocenici di Villalvernia (Ales-
sandria). I. Lamellibranchi. pp. 79-128, 4 figg., 10 tavv.

Volume XXI

I- CORNAGGIA CASTIGLIONI O. & CALEGARI G., 1978 - Corpus
delle pintaderas preistoriche italiane. Problematica, schede, iconogra-
fia. pp. 5-30, 6 figg., 13 tavv.

Il - PINNA G,, 1979 - Osteologia dello scheletro di Kritosaurus notabilis
(Lambe, 1914) del Museo Civico di Storia Naturale di Milano (Ornithi-

, schia Hadrosauridae). pp. 31-56, 3 figg., 9 tavv.

IN - BIANCOTTIA., 1981 - Geomorfologia dell'Alta Langa (Piemonte me-

ridionale). pp. 57-104, 28 figg., 12 tabb.,-I carta f.t.

Volume XXIII

I- GIACOBINI G., CALEGARI G. & PINNA G., 1982 - I resti umani fos-
sili della zona di Arena Po (Pavia). Descrizione e problematica di una
serie di reperti di probabile età paleolitica. pp. 5-44, 4 figg., 16 tav.

Il - POLUZZI A., 1982 - I Radiolari quaternari di un ambiente idrotermale
del Mar Tirreno. pp. 45-72, 3 figg.. 1 tab., 13 tavv.

III - ROSSI F., 1984 - Ammoniti del Kimmeridgiano superiore-Berriasiano
inferiore del Passo del Furlo (Appennino Umbro-Marchigiano). pp. 73-
138, 9 figg., 2 tabb., 8 tavv.

Volume XXIV
I- PINNA G,, 1984 - Osteologia di Drepanosaurus unguicaudatus, lepido-
sauro triassico del sottordine Lacertilia. pp. 5-28, 12 figg., 2 tavv.
Il - NOSOTTI S. e PINNA G., 1989 - Storia delle ricerche e degli studi
suî rettili Placodonti. Parte prima 1830-1902. pp. 29-86, 24 figg., 12
tavv.

Volume XXV i
I- CALEGARI G., 1989 - Le incisioni rupestri di Taouardei (Gao, Mali).
Problematica generale e repertorio iconografico. pp. 1-14, 9 figg., 24

tavv.

Il - PINNA G. & NOSOTTI S., 1989 - Anatomia, morfologia funzionale
e logia del rettile placodonte Psephoderma alpinum Meyer,
1858. pp. 15-50, 20 figg., 9 tavv. >

INI - CALDARA R., 1990 - Revisione Tassonomica delle specie paleartiche
del genere 7ychius Germar (Coleoptera Curculionidae). pp. 5/-2/8,

575 figg.

Volume XXVI va

I- PINNAG,, 1992 - Cyamodus hildegardis Peyer, 1931 (Reptilia, Placo-
dontia). pp. 1-21, 23 figg. :

II - CALEGARI G. a cura di, 1993 - L’arte e l’ambiente del Sahara preisto-
rico: dati e interpretazioni. pp. 25-556, 647 figg. ) na

INI - ANDRI E. e ROSSI F., 1993 - Genesi ed evoluzione di frangenti, cin-
ture, barriere ed atolli. Dalle stromatoliti alle comunità di scogliera mo-
dere. pp. 559-610, 49 figg.. 1 tav.

Volume XXVII : ,
I - PINNA G. and GHISELIN M. edited by, 1996 - Biology as History. N.
1. Systematic Biology as an Historical Science. pp. /-/33, 68 figs.
Il - LEONARDI C. e SASSI D. a cura di, 1997 - Studi geobotanici ed en-
tomofaunistici nel Parco Regionale del Monte Barro. pp. / 35-266, 122
figg. 23 tabb.

Volume XXVIII
1 - BANFI E. & GALASSO G., 1998 - La flora spontanea della città di
Milano alle soglie del terzo millennio e i suoi cambiamenti a partire dal
1700. pp. 267-388, 71 figg., 30 tabb.

Volume XXIX
I- CALEGARI G,, 1999 - L’arte rupestre dell’Eritrea. Repertorio ragionato
ed esegesi iconografica. pp. /-174, 268 figg.

Volume XXX

1- PEZZOTTA F. edited by, 2000 - Mineralogy and petrology of shallow
depth pegmatites. Paper from the First International Workshop. pp.
1-117, 30 figs., 19 tabs.

Il - PARISI B., FRANCHINO A. & BERTI A. con la collaborazione di PO-
TENZA B. & RUBINI D., 2000 - La Società Italiana di Scienze Natu-
rali 1855 - 2000. Percorsi storici. pp. /-163, 199 figg.

Ill - DE ANGELI A. & GARASSINO A,, 2002 - Galatheid, chirostylid and
porcellanid decapods (Crustacea, Decapoda, Anomura) from the Eoce-
ne and Oligocene of Vicenza (N Italy). pp. 1-31, 27 figs., 9 pls.

Volume XXXI

I- NOSOTTI S. & RIEPPELO., 2002 - The braincase of Placodus A gassiz,
1833 (Reptilia, Placodontia). pp. /-/8, 15 figs.

Il - MARTORELLI G., 2002 - Monografia illustrata degli uccelli di rapina
in Italia. (1895). Riedizione a cura di Fausto Barbagli. pp. /XX7 1-216,
[14] 46 figg., 4 tavv.

III - NOSOTTI S. & RIEPPELO., 2003 - Eusaurosphargis dalsassoi n. gen.
n. sp., a new, unusual diapsid reptile from the Middle Triassic of Besano
(Lombardy, N Italy). pp. 1-33, /9 figs., / tab., 3 pls.

Volume XXXII

I- ALESSANDRELLO A., BRACCHI G. & RIOU B., 2004 - Polychaete,
sipunculan and enteropneust worms from the Lower Callovian (Middle
Ties) of La Voulte-sur-Rh6ne (Ardèche, France). pp. 1-16, 9 figs.,

pl.

Il - RIEPPEL O. & HEAD J. J., 2004 - New specimens of the fossil snake
genus Eupodophis Rage & Escuillié, from Cenomanian (Late Creta-
ceous) of Lebanon. pp. 1-26, 13 figs., 1 tab.

INl - BRACCHI G. & ALESSANDRELLO A.., 2005 - Paleodiversity of the
free-living polychaetes (Annelida, Polychaeta) and description of new
taxa from the Upper Cretaceous Lagersrditten of Hagel, Hadjula and
Al-Namoura (Lebanon). pp. 1-48, 8 figs., 1 tab., 16 pls.

Volume XXXIII

I- BOESIA. & CARDI F. edited by, 2005 - Wildlife and plants in tradi-
tional and modern Tibet: conception, exploration and conservation. pp.
1-88, 30 figs., 9 tabs. ;

II - BANFI E., BRACCHI G., GALASSO . & ROMANI E., 2005 - Agro-
stologia Placentina. pp. 1-80, 7 figs., 1 tab.

INI - LIVI P. a cura di 2005 - I fondi speciali della Biblioteca del Museo
Civico di Storia Naturale di Milano. La raccolta di stampe antiche del
Centro Studi Archeologia Africana. pp. /-250, 389 figs.

Volume XXXIV

I- GARASSINO A. & SCHWEIGERT G., 2006 - The Upper Jurassic
Solnhofen decapod crustacean fauna: review of the types from old de-
scriptions. Part I. Infraorders Astacidea, Thalassinidea and Palinura. pp.
1-64, 12 figs., 20 pls.

Il - FUCHS D., 2006 - Morphology, taxonomy and diversity of vampyropod
Coleoids (Cephalopoda) from the Upper Cretaceous of Lebanon. pp.
1-28, 9 figs., 9 pls.

Ill - CALDWELL M. W., 2006 - A new species of Pontosaurus (Squamata,
Pythonomorpha) from the Upper Cretaceous of Lebanon and a phylo-
genetic analysis of Pythonomorpha. pp. 1-42, 18 figs., 1 pl.

Volume XXXV

I- DE ANGELI A. & GARASSINO A., 2006 - Catalog and bibliography of
the fossil Stomatopoda and Decapoda from Italy. pp. 1-95.

II - GARASSINO A., FELDMANN R.M. & TERUZZI G., edited by, 2007 -
3" Symposium on Mesozoic and Cenozoic Decapod Crustaceans. Mu-
seo di Storia Naturale di Milano May 23-25, 2007. pp. /-104, 38 figs.,
6 tabs.

II - NOSOTTI S., 2007 - Tanystropheus longobardicus (Reptilia, Proto-
saura): re-interpretations of the anatomy based on new specimens from
the Middle Triassic of Besano (Lombardy, N Italy). pp. 1-88, 67 figs..
4 pls., 9 tabs.

Volume XXXVI

I- GALASSO G., CHIOZZI G., AZUMA M. & BANFI E, a cura di,
2008 -

Le specie alloctone in Italia: censimenti, invasività e piani di azione.
Milano, 27-28 Novembre 2008. pp. 1-96.

Il - MAGANUCO S., STEYER I. S., PASINI G., BOULAY M., LOR-
RAIN S., BENETEAU A. & AUDITORE A., 2009 - An exquisite
specimen of Edingerella madagascariensis (Temnospondyli) from the
Lower Triassic of NW Madagascar; cranial anatomy, phylogeny, and
restorations. pp. 1-72, 32 figs., 1 tab., 4 appendix.

III - GARASSINO A,, 2009 - The thoracic sternum and spermatheca in
the extant genera of the family Homolidae De Haan, 1839 (Crustacea,
Decapoda, Brachyura). pp. 1-80, 2 figs., 18 pls., 1 tab.

Appendicular skeleton / Scheletro appendicolare

I-II

1-4

ac
actil

dorsal neural arch

arco neurale dorsale
diapophysis

diapofisi

dorsal rib

costola dorsale

epipophysis

epipofisi

gastralia

haemal arch (chevron)

arco emale

interspinal ligament attachment
inserzione dei legamenti interspinali
infrapostzygapophyseal fossa
fossa infrapostzigapofisaria
lateral gastralium

gastralium laterale
mediodorsal facet

faccetta mediodorsale

medial gastralium

gastralium mediale
medioventral facet

faccetta medioventrale

neural canal

canale neurale

neurocentral suture

sutura neurocentrale
neurocentral articular surface
superficie articolare neurocentrale
neural spine

spina neurale

posterior centrodiapophyseal lamina
lamina centrodiapofisaria posteriore
pneumatopore

pneumatoporo
postzygodiapophyseal lamina
lamina postzigodiapofisaria
postzygapophysis
postzigapofisi
paradiapophyseal lamina
lamina paradiapofisaria
prezygodiapophyseal lamina
lamina prezigodiapofisaria
prezygoepipophyseal lamina
lamina prezigoepipofisaria
prezygoparapophyseal lamina
lamina prezigoparapofisaria
prezygapophysis

prezigapofisi

sacral vertebra

vertebra sacrale

sacral centrum

centro sacrale

sacral neural arch

arco neurale sacrale

sacral rib

costola sacrale

transverse process

processo trasverso
intrapostzygapophyseal fossa
fossa infrapostzigapofisaria
intrapostzygapophyseal pneumatopore
pneumatoporo infrapostzigapofisario
tuberculum

tubercolo

wing-like expansion of the medioventral

facet
espansione ad ala della faccetta
medioventrale

first to third digit

dito dal primo al terzo

first to fourth phalanx

falangi dalla prima alla quarta
acromion

acetabular portion of ilium
porzione acetabolare dell’ileo

actis

bf
ca

cabil

ccil

clf

acetabular portion of ischium
porzione acetabolare dell’ischio
brevis fossa

carpals

carpali

caudal blade of ilium
lama caudale dell’ileo
cranial concavity of ilium
concavità craniale dell’ileo
fossa of collateral ligament
fossa del legamento collaterale
cnemial crest

cresta cnemiale

coracoid

coracoide

coracoid foramen

forame coracoideo
coracoid tubercle
tubercolo del coracoide
cranial blade of ilium
lama craniale dell’ileo
cranioventral process
processo cranioventrale,
distal carpals 1+2

carpali distali 1+2
deltopectoral crest

cresta deltopettorale
epicleideum

extensor pit

fossa dell’estensore
femur

femore

fibula

fibular condyle

condilo fibulare

flexor tubercle

tubercolo del flessore
furcula

symphysis of furcula
sinfisi della furcula
glenoid fossa

fossa glenoidea

greater trochanter
trocantere maggiore
humeral head

testa dell’omero

hooked process of ilium
processo uncinato dell’ileo
humerus

omero

hypocleideum

ilium

ileo

iliac process of ischium
processo iliaco dell’ischio
iliac process of pubis
processoviliaco del pube
internal condyle

condilo interno

ischium

ischio

ischial foot

piede ischiatico

ischial medial facet
faccetta mediale dell’ischio
ischial peduncle
peduncolo ischiatico
incisura tibialis

lateral epicondyle
epicondilo laterale

lesser trochanter
trocantere minore
metacarpal

metacarpale

medial epicondyle
epicondilo mediale
medial fossa of fibula
fossa mediale della fibula

l; SUS

Lan

Le Memorie sono disponibili presso la Segreteria della Società Italiana di Scienze 1
Museo Civico di Storia Naturale, Corso Venezia 55 - 20121 Ma
Pubblicazione disponibile al O cpr AI

.
4 1A
"L 6 0
n - : I, 4 ‘ , DI; 3 i. ì
c bi di * P ci i Li “
i ; Ò % E INDI MENdeio.
è é : VAS d
dù À - Chi a E A POME
fa i 4 Pe. ( hi TI Ca al > s e ni “4
: ‘ ” CRE "I ; fra i (0 »
È £ ” Lar te el 0 PER 1% ns Pia Ia Li