136 
ME, W. K. PAEKEE ON THE STRUCT UEE AND 
often not be exhibited until the cartilage lias in many cases been converted into bone. 
The skull of a Shark and that of a young Rodent (after the osseous centres are all 
formed) may be profitably compared together ; for it would seem that in the Mammalia 
alone does the skull perfect its segments — segments altogether so unlike those which 
form the rest of the axis, and which are most truly primordial*. 
The membranous interorbital space in the young Salmon is the equivalent of the 
“ fenestra” in the Bird, formed by retreat of the ethmo-presphenoidal cartilage from the 
trabecular crest. In Pelecanus onocrotalus both the prseoral arches behave in a similar 
manner; both coalesce in front of the elbowed part, and both of these double bars send 
upwards an azygous keel. I shall have to return to this comparison in describing the 
next stage. The trabecular carina only reaches to the converged part of the trabecula ; 
the ethmo-presphenoidal wedge only takes up the front third of the prosencephalic 
(anterior sphenoidal) region (Plate V. fig. 2). The cranial roof is thicker, and extends 
further over the middle cerebral lobe than in the last stage, and the band which 
unites the roof to the ear-capsule is broader and shorter. The alisplienoidal region 
is beginning to be walled-in by a growth of cartilage downwards from the band, and 
forwards and inwards from the ear-sac ; for the rest, this region is membranous, as in 
the Lacertilia. The Lizards have an orbito-splienoid mapped out by an outline of partly 
ossified cartilage, and the band at the top of this region answers, as far as it goes, to the 
very edge of the roof-cartilage of the young Salmon ; but there is no band running to the 
periotic region, and the only probable rudiment of an alisplienoid in the skulls in my 
collection is a small epiphysis at the antero-superior angle of the prootic in a Mexican 
Lizard ( Loemanctus longipes). In the young Salmon the anterior sphenoidal region, 
although w 7 ell roofed-in, has no cartilage in its side walls, as in the carinate Birds and 
“ Strutliionicke,” with the exception of Struthio. The “ fontanelle ” is still very large 
(fo.), yet the superoccipital cartilage ( s.o .) reaches to the junction of the anterior and 
posterior semicircular canals ( a.s.c ., p.s.c.). Plow large these and the other parts of the 
labyrinth are the figure shows, and also how the periotic cartilage fails to enclose the 
ampullae and most of the arch of the anterior and posterior canals, as well as the 
“ utriculus ” (lit.) and “ sacculus ” (sc.). The strong sheath of the notochord is now 
beginning to be ossified, and thus to lay the foundation of the “ basioccipital,” but of it 
alone; the posterior part of the investing mass” (iv.) forms additional substance for 
the completion of this quasi centrum. The roof-bones are not shown in this figure, 
except the “ superethmoidal lamina the premaxillary is also cut through ; and the long 
“ parasphenoid ” has now the “ vomer ” beneath its fore end, and to it are attached a 
number of teeth. 
A section (Plate V. fig. 3) immediately in front of the projecting “ ectoethmoidal 
wings” (prefrontal lobes) shows at this stage a very complete coalescence of the upgrowths 
* In all generalizations of this sort the AmpMoxus is left out of the question; when we obtain links that will 
in any way bind it to the Myxinoids, then we may begin to reason from a higher stand-point (see “ Erog’s Skull,” 
p. 202). 
