The avian relationships of Archaeopteryx are assessed in terms of the ‘stem-group’ concept. The avian stem-group is defined, and its constituents are identified and described. Phylogtnetic analysis of stem-group birds reveals that Archaeopteryx is no more closely related to modern birds than are several types of theropod dinosaurs, including lyrannosaurids and ornithomimids. Archaeopteryx is not an ancestral bird, nor is it an ‘ideal intermediate’ between reptiles and birds. There are no derived characters uniquely shared by Archaeopteryx and modern birds alone; consequently there is little justification for continuing to classify Archaeopteryx as a bird. Feathers are considered to be homologues or derivatives of epidermal scales (not of hairs); they probably originated as an insulating blanket in juvenile theropods, enabling them to match the activities o bigger animals, regardless of environmental temperature fluctuations. The furcula may have been present in various theropods, including allosaurids, caenagnathids (oviraptorids) and tyrannosaurids.
Several possible definitions of the class Aves are examined. It is concluded that the boundary between reptiles and birds is best placed at a pronounced ‘morphological gap’. This measure ensures that most animals conventionally regarded as ‘birds’ will be retained in the class Aves though Archaeopteryx would be transferred to the dinosaur suborder Theropoda. This definition also ensures that birds will be distinguished from reptiles by an extensive set of ostcological characters. The origin of the class Aves (as defined here) would probably coincide with the origin of avian flight; it is unlikely that Archaeopteryx can provide any direct evidence about the origin of modern avian flight, regardless of its locomotor abilities. The ‘morphological gap’ between reptiles and birds is not necessarily a deficiency of the fossil record; the ‘gap’ may be real evidence that birds originated by evolutionary saltation—and not by gradual stages. The origin of birds is not a problem to be equated with the origin of Archaeopteryx; it is a problem to be found in the ‘morphological gap’ that precedes the first appearance of volant birds. The concept of the ‘proavis’ has outlived any usefulness it might once have had, and should be abandoned.