ODONTORNITHES 651 
phrase of it as a “Swimming Ostrich.” About the same time it 
was found that the presence of teeth was a character apparently 
common to all ‘Cretaceous” Birds. The opinions expressed by 
Prof. Fiirbringer in the earlier portion of his great work need not 
be here adduced, since they were modified in the course of its pro- 
gress; but he finally declared (Untersuchungen, u. s. w. pp. 1543, 
1565, 1580) that the Odontolew were the ancestral relations of 
his ‘“ Colymbo-Podicipites,” with which they formed his Suborder 
“ Podicipitiformes,” while a similar view was taken in 1890 by 
Prof. D’Arcy Thompson (Stud. Mus. Dundee, No. 10). 
As to the Odontotormz Prof. Marsh has displayed commendable 
caution. On account of some similarity, the significance of which 
may or may not be important, he based his restoration of [chthyornis 
on Sterna (TERN), a fact which has led to exaggerated if not mis- 
taken views, for he was careful to state that Ichthyornis seemed to 
have points of resemblance to Ardea, Ciconia, Colymbus, Phalacro- 
corax and even to the Accipitres, while its posterior extremities alone 
indicated a structure similar to that of the Laridw# and Alcidz. In 
1893 Dr. Gadow (Thier-reich, Vogel, i. p. 119) suggested that the low 
characters of Jchthyornis shew it to be the beginning of the Carinatz. 
In 1891 the present writer (Cat. Foss. B. Br. Mus. pp. 200 et seqq.), 
while fully admitting both the Colymbine affinities of Hesperornis 
and the Larine resemblances of Jchthyornis, proposed to retain the 
term Odontornithes for a series of Birds ancestral to the modern series 
of toothless Carinatx, adopting (op. cit. p. 2) for the latter, but in a 
wider sense, Dr. Stejneger’s name (Stand. Nat. Hist. iv. p. 64) of 
Ewornithes. In addition to the presence of teeth, the extinct series 
differs from the Huornithes by the absence of union between the rami 
of the mandible, and between the distal ends of the ilium and ischium. 
“Whatever may be the ultimate verdict on these points of 
classification, it would seem probable that Hesperornis should be 
regarded as an offshoot from the same ancestral stock as the’ one 
from which the modern Colymbidx have originated; such ancestral 
stock being characterized by the presence of tceth, absence of 
ancylosis between the mandibular rami, and want of union 
between the spike-like patella and the upwardly-produced cnemial 
crest of the tibia. On the other hand, the abortion of the keel of 
the sternum, as well as the gencral Ratite modification of the 
scapular arch, are features peculiar to Hesperornis, and not common 
to the ancestral type; being, in fact, nearly analogous to those 
presented by Didus among the Columbe. 'The typical species of 
Hesperornis (H. regalis) was of large size, attaining a length of about 
six feet; while a second species (H. crassipes) was still more gigantic. 
Both were aquatic, and probably very similar in their general 
habits to the Divers. Probably more or less closely allied to this 
genus was the much smaller Colymbiform bird from the Cambridge 
