270 
TKIiTIAK^ VEiiTEBRATA OF THE FAYtjM. 
'J’urtlcs exce])t >Sj>har(/is, the only liviiiiJ- representative of the Athecte. The bone, in 
fact, 1 )elongs to tlie most highly specialised type of swimming humenis, to which 
Wicland * lias given the name parathalassic. The whole bone is strongly compressed 
dorsi-venti-ally. The liead (b.), so far as preserved, is strongly convex and somewhat 
triangular in outline. The nlnar crest (a.) projects further beyond the head than 
in Pscpliophorus scaldii or in Sphargis. Between the anterior thickening, which 
terminates on the head, and the posterior border the surface of the shaft is concave 
Text-fig. 87. 
Left humerus of Psephopliorus eoanms^ specimen ; dorsal and ventral views, 
n., ulnar crest ; h., head ; c., radial crest ; d., entocondyle. i nat. size. 
on the upper and lower faces of the bone, but the ventral concavity is deepest. The 
part of the bone above the radial process is considerably more elongated in proportion 
to its width than in FsepiiogPiorus scald'd and still more than in Sphargis. The radial 
prominence is very strongly developed, but it cannot be seen exactly what is the form 
of its ventral surface, which, however, is not divided into two or more separate knobs, 
as in Fsephophorus, and is continued on the ventral face of the bone obliquely back- 
wards, so that if the line of its direction were continued it would jiass through the end 
of the ulnar process ; in both Sphargis and Fsephoplioras the ridge is placed more 
transversely. The dorsal surface of the shaft opposite the base of the radial process 
bears a deep oval pit for the attachment of muscle; in Sphargis this seems to l)e 
represented by several smaller depressions. The distal end of the bone is incomplete, 
but, like the np})er end. it seems to have been less expanded than in Sphargis. There 
is no trace of any ectepicondylar groove or foramen. 
On tlie wliole, the bone resembles the humerus of Fsephophorus rather than that of 
* A trier. Journ. Sci. [4] vol. ix. 1900, p. 420, 
