DEVELOPMENT OE THE SKULL IN STURGEONS. 
173 
process is large and projecting. That is the only part not covered, on the outside, by 
the flattish but thick dentary bone (d.). On the inside (Plate 18, fig. 8, d .) it shows 
a large crescent of bone, where the rod becomes slenderer in the middle. Behind that 
part, on an apophysis of the cartilage, a small oblong second bone appears. This is 
the coronoid (cr.) ; its direction is backwards and a little downwards. The distal end 
of the rod is thick and massive, and is united to its fellow by ligamentous fibres. 
A small squarish suspensorial ray is seen behind the angle of the mandible (Plate 18, 
fig. 5, sp.r.) ; it is far below the usual plane for the spiracular cartilage of a Shark. 
The other half of the general roof of the mouth is fan-shaped (Plate 18, figs. 4 and 5, 
mt.pg'., mt.pg".), the narrow handle running in between the two pterygo-quadrates, 
and the outspread part growing round their convex hind margin. Thus this complex 
plate has two convex edges behind and two concave edges in front; then it runs to a 
sharp point between the symmetrical plates. The earliest pieces of this patchwork 
are still the largest; they are the main azygous plate and the main symmetrical plates 
[mt.pg'., mt.pg".) ; the single piece is like the bowl of a spoon, but longer; the other 
two are roughly three-cornered. Outside, between the latter and the quadrate region 
of the paired plate ( q.c.) there is a much smaller but tolerably constant piece on each 
side. Behind, between the three main patches, there are three or four on each side, 
inconstant in number and form ; and in front, in the interspace between the three 
main patches and the pterygo-quadrates, there is a patch or two on each side, and 
then a single row of four or five, lessening forwards. These patches are all set in one 
common web of fibrous tissue, so as to look, in a rough dissection, like one unpaired 
hard-palate plate. 
Half this complex palate of pterygo-quadrates and metapterygoids is nearly equal in 
length and width, but far inferior in thickness, to a single hyomandibular (Plate 18, 
fig. 5, lim.). That segment has no “serial homologue” either before or behind it, for the 
subdivision of the upper part of the visceral arches is different in the mandibular, hyoid, 
and branchial arches. The hyomandibular is not the uppermost segment of a normal 
visceral ( branchial ) arch, for there is no pharyngo-hyal in this type. Nor does it 
correspond to more than the upper two-thirds of an epi-branchial; the lower third is 
the separate symplectic ( sy .). Neither does it harmonise with the pier of the arch in 
front, which is segmented quite after another fashion ; there, the part answering to the 
symplectic is the pterygoid foregrowth of the suspensorium, whilst the part which 
should correspond to the head of the hyomandibular is partly continuous with the 
head of the pier of the opposite side, and largely broken up into a tesselated pave¬ 
ment of irregular segments. The upper “epi-hyal” (Jim.) is, above, a normal pha- 
langiform segment, but below and behind it expands into a huge pedate plate; the 
“ toe” below, is tied to the next segment, the “heel ” behind, carrying the “opercular,” 
and answering to the opercular process in Teleostei; this convex pedate slab is almost 
entirely unossified. 
