DEVELOPMENT OF THE SKULL OP THE COMMON EROO. 
143 
is by blunt points, which are curved inwards and then a little backwards ; the arches 
then gently curve forwards to the middle, from which part they turn backwards 
and inwards again. Thus, notwithstanding all the melting together of facial bars, 
sense-capsules, and cranial walls, in more advanced stages, nothing of the kind now exists; 
the after-blending is due to secondary connectives that fuse together the elementary organs. 
The first pair of these arches correspond to the “ palatal bands ” (Muller) of the larval 
Lamprey ( Ammocoetes ), but they indicate a much earlier stage ; for Muller’s figures 
(‘ Myxinoids,’ pi. 4, figs. 7-10, D) show a continuity of these bands both with each other 
in front, and with the “ investing mass ” of the notochord behind ; this will be illustrated 
in my “ third stage.” These bars, the “ trabeculae cranii ” of Rathke, are the shortest 
and the thickest of the series ; they retain their parallelism with the next pair as far 
downwards as the middle, and then turn forward and inward, lying like little beams 
beneath the totally distinct membranous cranium (see fig. 8, a transversely vertical view). 
The second of the series (first poststomal) is the rudiment of the mandibular arch ; but 
at present there is no segment freed from its lower end, answering to “ Meckel’s carti- 
lage.” The third, or second poststomal, is the hyomandibular and hyoid cornu in one 
undivided piece ; the fourth is the first branchial, and has a branchial “ bud ” upon it ; so 
also has the fifth, but not the sixth ; the seventh arch (sixth poststomal, fourth branchial) 
is not yet clearly differentiated. The horizontal views are very instructive : figs. 5-7 are 
of an embryo younger than that dissected from the outside (fig. 3) ; and fig. 9 is a cor- 
responding section to the lowermost of the earlier stage (fig. 7), in one somewhat older 
than that displayed in fig. 3. 
In fig. 5 the rods are cut through, by a somewhat oblique section, immediately 
below the cranial capsule ; the “ investing mass,” the rudimentary first vertebra (y. 1), 
and the enclosed notochord are displayed in this section, and also the shape of the 
mouth-chasm. Fig. 6 is a little lower down, and shows what fig. 5 does not, namely, 
the top of the sixth arch, or third branchial ; here the inner walls of the cheeks are seen 
to some extent. Lower down still (fig. 7) we see 'the floor of the mouth, with the rudi- 
mentary tongue (tg.) ; and behind the arches, only five of which could be clearly seen, 
we have the large yelk-mass (y.) with its emarginate front outline. Expecting to find 
the earliest condition of the “ trabeculte ” (first pair of rods) in the ammoccetine condition 
(that is, continuous with the “investing mass”), these sections cost me much thought, 
and these observations were repeated again and again. The fact is that the last of the 
arches comes nearest to the investing mass, and not the first , and it is wholly distinct 
from the axial structures. 
Thus we see that “ the branchial arches have the same morphological value as the 
hyoid, and the latter as the mandibular arc ” (Huxley, “ Croonian Lecture,” p. 53) ; and 
not only so, but the “ trabeculae cranii ” are merely the foremost of these arches, as 
both of us have for some time felt satisfied. 
In the earliest condition of my “first stage” (fig. 7), although these thick rods gra- 
dually decrease in size from before backwards, yet the first are twice as thick as the 
