OF THE MARSIPOBRANCH FISHES. 
425 
Cranio-facial skeleton of nearly metamorphosed young of Petromyzon marinus. 
(4 inches long.) 
This stage is still more instructive than the last, as the metamorphic state was not 
over,—the Petromyzine. skull was still in formation, and was caught in what may be 
called a Myxinoid stage (Plate 10, figs. 4, 5). Yet the suctorial (labial) cartilages were 
perfect (fig. 6) ; these have already been described, and the skeleton of the intervelar 
shelf (fig. 7), which was almost as much developed as that of the larger individuals 
of the same species (see Plate 18, fig. 7). 
If the reader will compare this minute chondrocranium with that of Myxine 
(Plate 10, figs. 1-8), with the larva (Plate 19, figs. 4, 5), and with that of the trans¬ 
forming Tadpole of the East Indian Bull Frog (Phil. Trans., 1881, Part I, Plate 4, 
figs. 8, 9), he will not fail to see its true morphological signification—both those skulls 
are twice the size of this, yet this belongs to the largest species of Lamprey.* 
Another skull which comes near to, and throws light upon this, is that of the 
transforming Tadpole of Pseudis (“ Batrachian Skull,” Part III., Plates 11, 12). Here 
the primary basal bands— para- and .pro-chordal (Plate 10, figs. 4, 5, iv., tr.), have their 
larval outline (see Plate 19, figs. 4, 5) perfectly distinguishable throughout. The 
chondrification creeps backwards, as in the early Tadpole’s skull, for here the hard 
cartilage is seen to end some distance in front of the hind margin of the ear-capsules 
(au.) ; the soft cartilage runs back along the spinal notochord for some distance. The 
sub-auditory wings of the parachordals are less than in the last stage and are very 
thin ; the space under the gasserian ganglion between the pedicle (pd.) and the capsule 
(au.) is much larger and more open than in the larger young (Plate 19, figs. 1—3). 
The basi-cranial fontanelle ( b.c.f. ) is now pointed in front— sagittate; and narrower 
than it will be in a few weeks. 
The trabeculae (tr.) in front of the pedicle (yd.) become half the size of the para¬ 
chordals, and then widen out, both in thickness and distance, to form the Ammocce- 
tine loop. The depth of the floor is made by the thickness of the trabeculae and its 
own concavity; and the bands are oval in section, the long axis of the oval looking 
upwards and inwards (see Plate 26, figs. 7-9, tr.). The interspace of the loop is 
accurately fitted with a cartilage, like the bowl of a spoon, but with two points, 
behind, instead of a handle ; this bifurcation gives the lanceolate form to the basi¬ 
cranial fontanelle (b.c.f). 
This thin shell-like median element is the “ posterior intertrabecula ” (p.i.tr.); 
although only transitorily distinct, and composed of hard cartilage, it is the true 
counterpart of the very distinct ladle-shaped, soft cartilage, seen in the adult Myxi- 
noids (same Plate, figs. 2, 3 , p.i.tr.). This element is more distinct, as a rule, in the 
* I stall describe the larval skull last (Plate 19, figs. 4 and 5) ; yet that simple platform of a 
cbondrocranium must be kept in sight, if the more advanced condition is to be understood. I have 
thought it better to take the metamorphosed skull first, as Anatomists are most familiar with it. 
