Polarornis is a controversial (see Mayr, 2002) genus of prehistoric bird. It contains a single species Polarornis gregorii known from incomplete remains of one individual found on Seymour Island, Antarctica, in rocks which are claimed to be of Late Cretaceous (López de Bertodano Formation, c. 76 mya) origin by supporters of the bird's antiquity, but may be no older than Eocene or about 50 mya.
The relationships of this taxon are unclear. It is not infrequently claimed to be an ancestor of modern loons (divers), but this is not necessarily correct as the material is not very diagnostic and the reconstructions take very much leeway to depict this bird as a primitive loon. What can be said with a reasonable degree of certainty is that it was in all likelihood aquatic and fed on fish and large invertebrates, probably being an ecological equivalent of loons, grebes, or the Cretaceous Hesperornithes of the Northern Hemisphere. Analysis of the structure of a femur (TTU P 9265) indicates that Polarornis was a flightless or near-flightless diving bird along the lines of the Hesperornithes or penguins, making a direct connection with modern looks highly unlikely (Chinsamy et al. 1998).
The theory has been proposed that Polarornis represents a Southern Hemisphere radiation of loon ancestors which filled the same ecological niche as the Hesperornithes and might even have prevented them from spreading southwards. Although conjectural, as clearly recognizable loons do not appear in the fossil record until the Late Eocene (some 35 mya), this theory is not entirely without merit: the Late Cretaceous genus Neogaeornis and Lonchodytes may be ancestral loons - indeed, the former is sometimes considered synonymous with Polarornis - which are known from South and North American locations, respectively. However, both are alternatively considered to be Hesperornithes or other seabirds; at any rate, there is no unequivocal proof that loons ever occurred in the Southern Hemisphere, and the overwhelming majority of evidence clearly points to a Western European origin of the order (see Loon for details).
It is notable that Eocene penguins from Seymour Island are somewhat intermediate between modern penguins and loons in the characters of their skull (Olson, 1985), and the least controversial interpretation of Polarornis is to consider it a basal neognathe bird of a lineage which later evolved into some sort(s) of seabirds like loons and/or penguins. The possibility that it is a non-neornithine cannot be ruled out completely at present.
Biogeography of avian survivorship of the Cretaceous–Tertiary extinction event coupled with new data on avian phylogeny (e.g. Fain & Houde 2004) suggests that survival of at least one, more probably some, southern Pacific lineage(s) of Neoaves is indeed likely. The possibility that Polarornis represents a valid Mesozoic taxon in some kind of basal relationship to any or all of such birds as plotopterids, cormorants, albatrosses, penguins, storks and/or loons is likely enough to warrant more research of the remains.