DEVELOPMENT OE THE SKULL IN THE SALMON. 
115 
lyriform afterwards ; the blades pass round the pituitary body, to which they are as 
nearly related at the beginning as they are in the Frog in my third stage (op. cit. 
Plate iv. fig. 9, tr.). There is no clear distinction at present between that layer of epi- 
blast which forms the skin and that which invests the rudiments of the brain as the pri- 
mordial or membranous cranium. Hence the position of the trabecula; (see Plate I. 
fig. 8, tr.) is between the epiblastic cerebral tube (imperfectly closed) and the palatal 
skin ( enderon ), and they are formed out of and enclosed in that thin tract of “ mesoblast ” 
which becomes differentiated not only into these trabecula, but also into the membranous 
cranium*. 
There is a great difference in the embryos at this early stage ; for while some have their 
mouth gaping and the angle of the mandible drawn back, others (see fig. 3) have a pecu- 
liar likeness to their sire , the old male Salmon, the rudiments of the lower jaw being very 
long and strongly hooked upwards ; these, however, are best for a contemplation of the 
morphological nature of the mouth. The mouth of the Frog-embryo caused me so much 
trouble that I almost despaired of classifying it (op. cit. p. 145, and note), but now I 
seem to “feel the light” from this translucent embryo. Looking at my third figure, and 
considering that the foremost arch is wrapped up in the succeeding folds of the face, then 
the seriality of the converging bars and intervening furrows will be understood. 
The facial clefts formed by dehiscence of the thinned interspaces of the rib-like 
thickenings of the face are not formed at the same time ; the first cleft, which I have dis- 
covered between the first and second preorals, does not become thoroughly distinct until 
after hatching. The first yjos^oral cleft is now visible (fig. 3), but it has lagged behind 
the second, for the cleft in front of the mandibular bar was formed first. 
That cleft is not one of a pair like the rest, but is double, and forms an azygous V-shaped 
opening, bounded in front by the second preoral, and behind by the first postoral bar ; this 
is the mouth (m). The palatines are primordial structures as distinct, even now, as the 
eyeballs which rest upon them ; they bound the opening gape above, and reach in front to 
the nasal sac, and behind to the auditory capsule, overlapping there the remarkable super- 
orbital band (s.ob.). Now, in conformity with their long suppression and late appear- 
ance in the Frog, even here in the Salmon their differentiation into hyaline cartilage 
is very tardy ; in my fourth stage the rod will be seen to enclose a protoplasmic pith, 
and in the fifth their apex will have melted into the fore edge of the next arch behind f. 
The section (Plate I. fig. 8, in) will show what space is forming for the mouth and 
* In the longitudinal section (Plate I. fig. 8) the trabecula (tr.) is coloured, as though it were showing through 
the palatal skin, for it could not be shown in a section of the skin, being at a little distance from the exact mid 
line ; the same is partly true of the next bar, pterygo-palatine, hut the rest are made naked at their very ends. 
t Here let me warn the reader against relying on terms of mere local relation, such as prefrontal, postfrontal, 
and the like : the term s«&ocular is good for immediate use in description ; hut at first the trabecula is the 
“ subocular ” in the Erog-embryo ; then in the Tadpole (compare “ Frog’s Skull,” Plate iv. fig. 1, tr., with Plate v. 
fig. 1) it is the mandibular pier (1st postoral) which is the subocular bar; here, in the Salmon, the subocular 
bar is the rudiment of the “ pterygo-palatine arcade.” We must not rest until our knowledge is put upon a 
sound morphological basis. 
