18 ON THE STRUCTURE OF EUSTHENOPTERON 
in the specimen figured on Plate 9, fig. 4), near the anterior 
margin of the palatine, which communicates with the mouth and 
seems to be a passage for the internal nares. 
Behind the parasphenoid there is a large vacuity bounded on 
either side by two, more or less triangular bones, each of which 
is perforated by a foramen. Their position and the presence of 
this foramen suggest their homology with the prootics of modern 
fishes, and I have so named them. The outer and hinder margins 
of the prootics articulate with a small quadrangular bone (shown 
in side view on Plate 9, fig. 3), which extends nearly vertically 
downward from the roof of the skull in the region of the supra- 
temporal. ‘They are also connected behind with a median shield- 
shaped bone which, it seems to me, represents the basi-occipital. 
I have figured this portion of the skull in ventral aspect on Plate 
14, figs. 4 and 5. Fig. 2 is an upper view of the same specimen 
showing the relation of these bones to the parietal region, while 
fig. 3 on the same plate shows it from the rear. 
Above the basi-occipital and extending to the roof of the skull 
is a pair of curved bones similar in every respect to the lateral 
elements of the vertebral centra, except that they are wider; and 
which I have called the exoccipitals. Like the centra they each 
bear on their outer surfaces a wide vertical sulcus. Each of them 
is perforated by a small foramen and each is in contact at its for- 
ward margin with a dumb-bell shaped bone which extends upward 
and outward to the roof of the skull. I am unable to offer any 
opinion as to the homology of this latter pair of bones except that 
they resemble in shape and position the paroccipitals in Trimer- 
orhachis as figured by Broom (1913, p. 572, fig. 6) and might 
therefore represent the opisthotics of fishes; but indeed, all of the 
bones of the posterior half of the brain case are so vastly different 
from those of any known fish or amphibian as to defy positive 
identification. 
Besides the bones I have described there are two other paired 
elements to be seen in the specimen shown on Plate 14, figs. 4 
and 5. One of them connects the prootic with the inner margin 
of the dumb-bell shaped bone, the other connects the outer margin 
of the last mentioned bone to the quadrangular element below the 
supratemporal just in front of which seems to have been the 
attachment for the hyomandibular. 
I shall not at.this time discuss the significance of the fact that 
Yh yh 
