DEVELOPMENT OE THE SKULL IN THE SALMON. 
121 
cleft from apex to base into a broad anterior and a narrow posterior band. The round 
distal piece belongs rather to the extremity of the former than to that of the latter ; 
thus the front division is connected with the azygous £: lingual ” or “ basihyal” piece by 
this distal segment, the “ hypohyal.” It is useless looking to the Tadpole for an expla- 
nation of this; and as we shall soon arrive at a “ Ganoid” stage, we may do well to look 
among the “ Plagiostomes ” — that is, we must search a lower morphological level. 
In the Brazilian Torpedo (Narcine) * there are two bars in this region, the foremost, 
however, being slender and the other broad (see Muller’s ‘Myxinoids,’ pi. 5. figs. 3 & 4, 
ci.d.). In my last paper (p= 195) I ventured to call the foremost bar a “ metapterygoid,” 
much wondering why it rested upon the broad hyoid suspensorium. I retract ; it is 
quite behind the mandibular region, and the two together are surely the split-up second 
postoral, the anterior a feeble “ hyo-mandibular,” and the posterior a massive “cerato- 
hyal;” there is no “hypohyal,” as far as I know, until we get to the “Ganoid” Sturgeon. 
In the higher groups we have long been pestered with two cartilaginous and bony tracts 
on each side in the substance of the second postoral arch ; hence the misunderstanding 
of the relations of the “incus;” this element may notv be said to be safely fixed in its 
own place as the “ antero-superior segment” of the split-up hyoid arch, and the mam- 
malian representative of the Fish's “ hyo-mandibular.” 
The “basihyal” (Plate II. fig. 3, gh.) has now become a thick lozenge-shaped mass of 
cartilage, strongly wedged betwmen the “ hypohyals.” This preparation, which has been 
flattened out for display, shows considerable modification in the form of the mandibular 
bars ; they are more strongly sigmoid, and the “ orbitar process,” above the abrupt bend, 
is very distinct ; its most definite reappearance is in Chelonians and Birds. The thick 
“ subocular ” flap is acquiring a more solid palato-pterygoid pith ; towards it, in front, 
the still small and distinct trabeculae ( tr .) are sending a small connective lobe, the 
first rudiment of the “ ectoethmoid ” with its “ superpalatal ” facet. The same 
stage is illustrated by the figure of an uncompressed specimen (Plate II. fig. 4), from 
which the postoral part of the face has been removed to display the auditory sacs from 
below. The oldest specimen of the first stage (Plate II. fig. 2) may illustrate this, 
which is but little older ; here (fig. 4) the “ notochord ” is twisted to the left side 
and projects beyond the ear-shaped flaps that form the end of the “ investing mass.” 
Behind, these moieties contract, first suddenly and then gradually, growing to a point as 
they form the “ basioccipital ” region. The primordial “ fenestra ovalis ” notches both 
the prootic cartilage and the investing mass, and with the help of the two secondary 
isthmuses a very elegant circular window is left ; not through it, but a little externally, 
the otolith may be seen shining through the cartilage. The anterior, horizontal, and pos- 
terior canals have bulged out the cartilage in their respective regions ; above, there is still a 
fenestra over the posterior canal ; this is not closed for some time to come (see next stage, 
fig. 7, ep.). Between the investing mass and the ear-sac in front there is a deep notch ; 
into this the trabecular apex is creeping. Another head at this stage has supplied me 
* Professor Gesex'bave in the recently published third part of his 4 Untersuchungen.’ 
R 2 
