Up: The dino-bird link
Archaeopteryx, the first bird, has feathers exactly identical to those of modern birds, hence it is closely related to them. It is linked to dinosaurs too, as shown by its nearly identical skeleton to that of small theropods (including the famous, recently identified 'sickle-claw'). The main argument against a dino-bird link was the lack of clavicles in dinosaurs, but there now are dinosaurs known with a developed wishbone.
See also All about Archaeopteryx at Talk.Origins
and the Archaeopteryx resource pages.
Dinosaurian features of Archaeopteryx
The skeleton of Archaeopteryx is nearly identical to that of small theropods, particularly deinonychosaurs such as Velociraptor, Deinonychus... Let us simply recall the numerous teeth, the long tail, the clawed fingers, and the dinosaur-like legs. In fact, the link of Archaeopteryx to dinosaurs is even clearer than its link to modern birds (though its feathers are exactly identical to modern birds').
Rather than a complete list of common anatomical features, here are the most striking facts:
An Archaeopteryx skeleton on which the feathers were not clearly visible was for twenty years misidentified as a Compsognathus fossil, though nearly complete and well-preserved.
A recent study of Archaeopteryx showed that it had the 'sickle-claw' (hyperextendable second toe with a very curved claw) which is one of the most distinctive features of deinonychosaurs.
Deinonychosaurs, unlike other theropods, have the backward-pointing pubis which is characteristic of modern birds.
Some dinosaurs are now known to have had a furcula (see below).
The arm of Archaeopteryx, i.e. its 'wing', is identical to that of deinonychosaurs, though proportionally bigger, but quite far from modern birds'.
Here are two more complete lists of dinosaurian characters of Archaeopteryx : one by Thomas Holtz and another one by G. Olshevsky at JDP.
There had been claims that a more primitive bird, Protoavis from the Triassic, was found, and therefore provided evidence that birds arose as early as dinosaurs and could not be descended from them. This fossil is very controversial and may even be a chimera.
Archaeopteryx is clearly related to birds, as shown by its feathers, and some other, less obvious skeletal features. It had been claimed that the feathers impressions on the fossils were a mystification, but 1) there are now 7 known Archaeopteryx skeletons, with feathers; 2) the feathers on the fossil and on its counter-impression in stone correspond exactly to each other; 3) there are other skeletal features that are intermediate between dinosaurs and birds; 4) there are other known birds, more modern-looking, with developed teeth; 5) there are other feathered dinosaurs.
The feathers of Archaeopteryx are exactly identical to those of modern birds in shape and disposition. In particular, they were asymmetrical, an adaptation clearly designed for flight.
Following Archaeopteryx, there is a sequence of fossils leading to modern birds: we can follow the reduction of the tail and teeth, the apparition of the carina (keel) for a more powerful flight, and the diversification of the group.
The wishbone debate
The furcula (or wishbone) is a bone resulting from the fusion of the two clavicles, which is necessary for flapping. Dinosaurs have long been thought not to have been able to give rise to birds since they completely lost the clavicles their ancestors had. Hence birds would have originated among these ancestors rather than among dinosaurs themselves.
Actually, the most primitive dinosaurs lack a furcula. But both new finds and re-examination of old fossil material show that lots of theropods had a well-developed furcula: these include some coelophysids (which are very primitive theropods), allosaurs, tyrannosaurs, troodonts, oviraptors, and of course, deinonychosaurs.
Three problems, and answers
The thecodont or crocodilian ancestry hypothesis. The hypothesis opposite to the dinosaurian origin of birds emphasised on thecodonts or crocodiles, which are close relatives to dinosaurs. For thecodonts, it was mainly motivated by the supposed loss of the clavicles in dinosaurs; actually, no group of thecodonts shows specific affinities with birds, whereas some dinosaurs really do. Crocodilians share a few similarities with birds, especially in the skull; but the skull of early crocodiles is already highly specialised, and the few common features seem not to come from a recent common ancestor.
The bird lung system is particular among all animals since the air goes only one way in the lungs, while additional air sacs continuously pump the air to the lungs. Some have considered this is difficult to evolve from dinosaurs, but it is exactly as difficult as from any other animal.
The bird digits debate. It has been claimed that the three-fingered bird hand kept fingers II, III and IV, while theropods kept I, II, III. The reasons why birds should have kept II, III, and IV is that 'generally', from embryological studies, animals which lose two fingers lose I and V. This should apply to theropods as well as to birds. Bird embryos already have only four fingers even in their earliest development stages, hence making the decision impossible from embryology.
Up: The dino-bird link