Burke & Feduccia Science 278:666

[augwhite@neosoft.com (Toby White)]

Summary: The theropod hand is undoubtedly I-II-III since IV & V are present
in reduced form.  _Archaeopteryx_ has been assumed to be I-II-III because
the phalangeal formula is the same as theropods.  However, this is a plastic
character and is not the same as in modern avians.  The authors assert that
hand development is very similar throughout the amniotes.  IV develops first
(the "primary axis") as a cartilagenous process from the ulna.  V develops
from the base of IV.  I-II-III develop (in the reverse order) as a "digital
arch".  Using the "primary axis" as a marker for IV, the authors show
preparations of embryonic hands from alligator, chicken and turtle which
indicate that the chicken manus is made up of II-III-IV, with a highly
reduced V being present.   They also assert an analogy between the
development of the hand and foot and show embryonic foot development in the
chicken which has all five members present, with II-III-IV predominating.
Finally, the authors asert that this pattern is highly conserved and that I
and V are always the first to be lost.  The theropod I-II-III pattern is
quite unusual and marks a significant departure from the base lineage.
Because chickens show the more typical tetrapod pattern, the authors believe
that birds are not descended from therapods.

Comments: GSP's long post just came in, and I expect it contains a better
critique than I can give.  A few comments only.  The authors say a good bit
about the association of IV and V with the ulnar process, but say nothing
about the derivation of the digital arch.  I'd like to see that worked out
since, from the pictures, its not at all clear that this is consistent.  The
pictures are quite convincing on IV, but unless you can  show derivation,
you're not doing embryology.  Of course, the central assumption is that the
primary axis is IV.  If that holds up, the paper is pretty strong.

  --Toby White