atic because acipenseriform larvae and juveniles
possess jaw teeth. Figure 20 shows the oral region of
a Polyodon larva, with two rows of conical teeth
erupting in the upper jaw and a single row in the
lower jaw (see also Bemis & Grande 1992). In large
adultPolyodon, however, (>25 kg) teeth are not vis-
ible from the surface of the bone, and sections show
that they are completely embedded in the jaw. The
pattern of ontogenetic tooth loss is different in stur-
geons: larvae have teeth, but the teeth and their at-
tachment bones are absent in adults, and are gener-
ally considered to be shed during growth. (The bit-
ing surfaces of adult sturgeons are composed of
thick collagenous pads, Nelson 1969.) Thus, this
character is more complicated than usually stated,
and available data for †Peipiaosteusand †Chon-
drosteusare inconclusive concerning the mode of
ontogenetic tooth loss.
Gardiner & Schaeffer (1989) also noted that aci-
penseriform scales are devoid of normal peg and
socket articulations, except on the caudal lobe of
†Chondrosteus. By analogy with phylogenetic
changes in scales of dipnoans (Bemis 1984), we sus-
pect that the loss of peg and socket articulations is
related to the general phenomenon of loss of scala-
tion in acipenseriforms (character 3 above), and
that both might be interpreted as paedomorphic.
Similarly, Gardiner & Schaeffer (1989) regarded
the absence of ganoin from fin rays as an acipenseri-
form character: this also could result from paedo-
morphosis (see Bemis 1984 for discussion of phylo-
genetic loss of ornament in dipnoans). There are
typically a few ganoid tubercles along the leading
edge of the pectoral fin in, for example,Acipenser
oxyrhynchus, but these do not extend into the area
of the fin membrane proper. Both of these putative
characters deserve study.
Discussion of clade Acipenseriformes
Based on the taxa and characters surveyed, there is
little doubt concerning monophyly of Acipenseri-
formes. Questions concerning sister-group rela-
tionships of the two Mesozoic families †Chondros-
teidae and †Peipiaosteidae arc discussed below, but
on the basis of available evidence we cannot reach
any conclusion and thus Figure 17 shows †Peipiaos-
teidae, †Chondrosteidae and Acipenseroidei as anunresolved trichotomy (see Crande & Bcniis 1996
for new information on this trichotomy).Characters of †PeipiaosteidaeCharacters of this group are the subject of another
study (Crande & Bemis 1996). and at present, we do
not accept any of the characters proposed by earlier
workers.Putative and problematic characters of † Peipiaostei-
dae
A character often cited as evidence of monophyly
of †Peipiaosteusis paired lateral line scales (Liu &
Zhou 1965, Zhou 1992). We think this is probably a
misinterpretation, for in several of the figured spec-
imens, the right and left trunk canals are preserved
as small tubular ossicles in close proximity to each
other as we have round for taxa such as †Cross-
opholis(Grande & Bemis 1991).
Zhou (1992) noted several characters may be syn-
apomorphies of †Peipiaosteus and †Stichopterus.
For example, the parasphenoid of†Peipiaosteushas
a broad anterior process, but this character has not
been fully assessed in †Stichopterus. In view of the
reported differences between these two genera
(Zhou 1992), it is important to make additional
specimen based study and comparison before ac-
cepting them.
†Peipiaosteus shares with Polyodontidae and
Acipenseridae the presence of small, tripartite den-
ticles on the shoulder girdle (Liu & Zhou 1965). Al-
though this character is presently unknown in †Sti-
chopterusand †Chondrosteus, it may be a synapo-
morphy of Polyodontidae, Acipenseridae and †Pei-
piaosteidae.
Liu & Zhou (1965) also noted that in †Peipiaos-
teidae, Acipenseridae and Polyodontidae, the pel-
vic fins originate anterior to the position of origin of
the dorsal fin, whereas in †Chondrosteus,the pelvic
fins originate opposite the dorsal fin. Again, this
may be a synapomorphy of Polyodontidae, Acipen-
seridae and †Peipiaosteidae.
Finally, †Peipiaosteusshares with Polyodontids
the presence of branchiostegals (there is only a sin-
gle branchiostegal in paddlefishes) with branched