Traditionally, the bird order Apodiformes contained three living families: the swifts, Apodidae, the tree swifts, Hemiprocnidae, and the hummingbirds, Trochilidae. In the Sibley-Ahlquist taxonomy, hummingbirds are separated from the Apodiformes as a new order, Trochiliformes. As their name suggests ("footless") their legs have limited function aside from perching. However, the hummingbirds, swifts and crested swifts do all share anatomical similarities to one another, and with their probable closest known living relatives, the owlet-nightjars (Mayr, 2002). The evolution of a short, stout humerus has provided the hummingbird with ideal wings for hovering, and the owlet-nightjars are apparently convergent with the closely related Caprimulgiformes, which form a cladeCypselomorphae with the Apodiformes (Mayr, 2002). Apodiformes have evolved on the Northern Hemisphere. Eocypselus, a primitive genus known from Late Paleocene or Early Eocene of north-central Europe is somewhat difficult to assign; Dyke et al. (2004) consider it a primitive hemiprocnid[1]. but most researchers believe that at present this genus cannot be unequivocally assigned to the Apodiformes or the Caprimulgiformes. The Early Eocene Primapus, found in England, is similar to both a primitive swift and the aegialornithids, which are in some aspects intermediate between swifts and owlet-nightjars. Fossil swifts are known to have existed during that time in Europe[2], and aegialornithids were possibly present in North America[3]. By the late Eocene (around 35 MYA), primitive hummingbirds started to diverge from the related jungornithids; the Middle Eocene Parargornis (Messel, Germany) and the Late Eocene Argornis, found in today's southernmost Russia, belong to this lineage. Cypselavus (Late Eocene - Early Oligocene of Quercy, France) was either a primitive hemiprocnid or an aegialornithid. The placement of the Aegialornithidae is not quite clear. Various analyses place them sufficiently close to the Apodiformes to be included here, or into the unique owlet-nightjar lineage in the Cypselomorphae. ORDER APODIFORMES
Dyke, Gareth J.; Waterhouse, David M. & Kristoffersen, Anette M. (2004): Three new fossil landbirds from the early Paleogene of Denmark. Bulletin of the Geological Society of Denmark51: 47–56. PDF fulltext
Mayr, Gerald (2002): Osteological evidence for paraphyly of the avian order Caprimulgiformes (nightjars and allies). Journal für Ornithologie143: 82–97. PDF fulltext
Mayr, Gerald (2003): Phylogeny of early tertiary swifts and hummingbirds (Aves: Apodiformes). Auk120(1): 145–151. DOI: DOI: 10.1642/0004-8038(2003)120[0145:POETSA]2.0.CO;2 PDF fulltext
Footnotes
^ This would suggest that the major apodiform lineages diverged already shortly after the K–T boundary. While this may well be so, the perching adaptation of the foot of Eocypselus on which they rest their case may just as well be a symplesiomorphy.
^ At that time, most of Europe had a humid, subtropical climate, possibly comparable to southern China. For a map of Early-Middle Eocene Earth, see here; note that both Caucasus mountains and Alps did not exist yet.
^ Somewhat doubtful, as this is difficult to reconcile with the other aegialornithid fossils and Primapus.
