It was during the dinosaur-era that the first birds appear in the fossil record, notably Archaeopteryx from Jurassic strata in Bavaria. Because it had fully developed feathers, including primaries and secondaries which were arranged on its wings in the same way as modern birds, Archaeopteryx is recognized and classified as a bird. Its skeleton, however, possessed a number of reptilian features. These included a long bony tail, claws on the digits of its fore -- as well as hind limbs, and teeth; so it is seen as a link -- showing that birds evolved from reptiles. It is widely held that birds arose from a group of dinosaurs known as the theropods. Features of the latter are their overall modest size, their bipedal stature, the long and clawed hind legs, and the three digits on their forelimb. Further, some of the theropods possessed hollow bones, and some later ones lacked teeth, with their mouth somewhat resembling a beak.
Archaeopteryx A legitimate transition?
Despite similarities between theropods and birds, there are substantial differences which very clearly militate against an ancestral relationship. Firstly, contrary to what might be predicted, the theropods were Saurischian, not Ornithischian, i.e. they had the distinctive lizard-like pelvis rather than a bird-like one. In addition, although both theropods and birds have only three digits in their forelimbs, they are different ones: theropods have the first to third of the usual five digits, whereas birds have the second to fourth. One particularly important difference is that the theropods -- whilst having long and well-developed hind limbs -- had very small forelimbs. Such short limbs were hardly credible precursors for the extended and well developed fore-limbs of birds, including those of Archaeopteryx, which must support their wings. Also, typical of running animals, they lacked clavicles which are present in Archaeopteryx and fused to form the familiar avian wishbone (furcula) which contributes to support of the wing. Finally, Archaeopteryx had a significantly larger braincase, reflecting the need for additional sensory input and motor control required for flying.
One of the major issues with reptile-to-bird evolution is that avian lungs are completely different from those of a reptile. No models have ever been presented that demonstrate how reptilian lungs could evolve into avian lungs and still serve their intended purpose.
It might also be noted that birds have a unique arrangement for distributing air around their bodies -- to support the high metabolism associated with flying. But there is nothing to indicate that any reptile group had such a system or even anything that could be considered a precursor to it. In fact, some scientists have concluded that the theropods had a crocodilian mode of breathing, and that to convert from this to the bird system would mean going through a non-viable intermediate stage.
A further important point is that the most bird-like theropods did not arise until late in the Cretaceous period -- about 70 million years after Archaeopteryx, so could not have been its ancestor. Indeed, they come in even later than various groups of Cretaceous birds which are closer to modern birds than to Archaeopteryx.
One final point which is worth mentioning is the problem of how powered flight started. Did wings develop by gliding down from trees or by running fast along the ground? There remains no convincing evidence either way, and indeed some evidence to the contrary either way. The point is that Archaeopteryx was a fully fledged bird, capable of flight, not an intermediate running or glider.
While it may be fair to conclude that Archaeopteryx was of intermediate structure (in the sense that it was a bird which shared a handful of features in common with reptiles), it cannot be regarded as transitional in terms of phylogeny. As with so many other classes of animal, the birds arise in the fossil record without any clear link to earlier creatures.