124 
ME. \\\ Iv. PARKEE OX THE STRUCTURE AXD 
of the whole matter will appear to him. As to the languid, feeble, tardy-growing 44 sub- 
ocular ” bands, these are not really cliondrified in this the third stage ; they are, however, 
larger. But the trabeculae have grown immensely (Plate II. figs. 6 & 7, tr.). Looking 
at them in this stage, before they have done more in relation to the cranium than to marry 
themselves to the “investing mass,” we shall see how facial arches must be modified that 
have to grow in a directly horizontal manner, adapting their apex to some convenient 
fastening-point behind. Save in the lack of segmentation of the out-bowecl part, these 
trabeculae at this stage are very exactly like the next pair (the palato-pterygoids) of the 
Pelecanidae, of Sulci and Pludacrocorax at present, and of Pelecanus itself, as soon as 
they have combined and sent upwards a common crest ; the divergence of the ptery- 
goids in these and other Birds is the true morphological equivalent of the divergence 
of the trabecular apices outside the pituitary space. 
We have seen already that the trabeculse at a very early stage (Plate I. fig. 5, tr.) 
have their leafy ends bilobate ; the outer lobe is the first morphological germ of the 
“ethmo-palatine ;” it, in reality, is a conjugational “stolon” from the minute trabecular 
stem. Already, in a very few days, the trabecula: have grown tenfold ; they have also 
articulated, as facial arches are wont to articulate, with their successor bars ; and now we 
have the junction of the ascending part of the palatine with the 44 lateral ethmoid,” 
never to be seen wanting again in our long ascent to Man. Underneath (fig. 6, tr.) the 
ethmoidal region is transversely ribbed : these ribs grow out afterwards as the pedicles 
(the serial counterparts of the 44 basipterygoid processes”), to which the palatines are 
suspended. In front the other lobe meets its fellow ; it is thick, and the two diverge at 
a large angle ; they together form the rudiments of the nasal septum and the “ subnasal 
laminae.” Above (fig. 7, tr.) the 44 ectoetlunoiclal” w T all is quite soft; it separates the 
nose from the eye : the whole of the cerebral roof is also quite uncartilaginous. 
The original membranous space below the primordial ear-opening is now a small 
fenestra over the posterior semicircular canal (fig. 7, ejy.) ; the 44 fenestra” in the periotic 
floor (fig. 6, f-s.o) is less; moreover, the whole capsule, now well united to the investing 
mass, has developed considerably since the last stage. Now the occipital arch (b. o., e.o.) 
can be seen behind the auditory capsules ; but the two divisions have not united above ; 
they are nearer, however, than the figure (7) would indicate, as it is a little outspread 
for display. 
The branchial arches do not grow at the same rate in different individuals; they are 
more differentiated in the oldest of my first stage (Plate II. fig. 2, hr.) than in those 
examined in this, the third. Plere (Plate II. fig. 8) these elegant blunt hooks have 
coalesced by their lower straight end, and the coalesced part is turned backwards, 
swollen, and half cut off as a median piece. This is similar to what I found in the 
formation of the sternum in the embryo ox ( 4 Shoulder-girdle and Sternum,’ plate xxix. 
fig. 1), where the median piece is made out of the ends of the lateral pieces. In the 
hyoid arch of the Salmon (Plate I. fig. 7, g.h.) the median piece is truly and primarily 
azygous ; here, in the branchial arches, it is a secondary azygous element, made up of 
the coalescing ends of a pair of rods. As yet there is no basal piece behind the third 
