4 
Though the general form of the posterior face of the fossil is similar to 
that of Pelecanus , the crests must have been quite different in arrangement. 
In the pelicans the inner crest of the hypotarsus is long, and is supported 
on a comparatively broad, strong base. The outer crest is much reduced, 
as it is only slightly more than one-third as long as the inner, and rises 
only slightly above the centre of its stronger brother. In Cyphornis , as 
has been indicated, the crests are broken away, but from the surfaces left 
we may adduce that the two parts of the hypotarsus were more or less 
similar in length, with the inner supported on a comparatively narrow base, 
and the outer much broader and heavier. The condition is strongly 
reminiscent of the arrangement in the Sulidae where the outer crest is 
double, and though only a little more than half as high is about three-fourths 
the length of the inner. The transverse groove above the proximal end of 
the hypotarsus, so well marked in Cyphornis, though very faintly developed 
in the pelicans, is also indicated in the Sulidae. In the cormorant- 
anhinga branch of the totipalmate birds the inner crest of the hypotarsus 
is high and blade-like with a narrow base, and the outer crest is greatly 
reduced. 
We find then that Cyphornis magnus though strongly pelican-like 
in its characters shows certain affinities to the Sulidae and more distantly 
to the Phalacrocoracidae and Anhingidae. Although similar to the Peleca- 
nidae, it is so different as to warrant its segregation in a distinct family, 
to be known as the Cyphornithidae. 
In this connexion there must be considered also Palaeochendides 
mioceanus Shufeldt 1 described from the lower end of a left femur, from 
Miocene deposits near the source of Stono river, South Carolina. Though 
Dr. Shufeldt considered this an anserine bird I have shown 2 that it is 
properly a species of the order Pelecaniformes (known formerly as the 
Steganopodes). After careful study of the type my conclusion as to the 
affinities of Palaeochendides mioceanus in the paper just cited (page 557) 
was as follows: 
“Should more of the skeleton become known, it may eventually be placed in a 
separate family. If we may venture to base theory on this one fragment, Palaeochendides 
was a pelican-like bird somewhat larger than Pelecanus erythrorhynchos or P. onocrotalus, 
as the portion of the femur representing it seems to indicate that the bone in its entirety 
was somewhat larger and heavier than the femur in these two species. In its appearance 
this bone seems, too, to show certain resemblances to the Sulidae and remotely to the 
Anhingidae and Phalacrocoracidae. Hence, while Palaeochendides will stand as a milepost 
in the line of descent of the pelicans, it brings down to us suggestions of generalized 
development indicating ancient relationships of pelicans to gannets and more remotely 
to the cormorant-anhinga branch of the totipalmates.” 
The affinities of Palaeochendides mioceanus are thus so similar to 
those adduced for Cyphornis that I have no hesitance in including it in 
the family Cyphornithidae. 
The arrangement will be as follows: 
Family, Cyphornithidae 
Genus, Cyphornis Cope 
Cyphornis magnus Cope, Jour, Acad. Nat. Sei., Philadelphia, ser. 2, vol. 9, p. 451 (May 31, 
1894). (Tertiary, probably Miocene of Vancouver inland.) 
Genus, Palaeochendides Shufeldt 
Palaeochendides mioceanus Shufeldt, Geol. Mag., N.S., dec. 6, vol. 3, p. 347, PI. 15 (August, 
1916). (Miocene of South Carolina.) 
1 GeoL Mag., N.S., dec. 4, vol. 3, p. 347, PL 15 (August, 1916). 
* Jour. Geol., vol. 25. pp. 555-557, fig. 1 (Scptember-October, 1917). 
