OF THE MARSIPOBRANCH FISHES. 
377 
vertebral arches. Moreover, the theca vertehralis is but little enlarged where it passes, 
insensibly, into the dura mater. 
The huge notochord (Plates 9 and 10, figs. 1, 2, nc.), with its merely membranous 
sheath, suddenly ends in a conical form between the ear capsules, its thick sheath 
becoming still thicker at the end, and the vacuoles and membranous bands ceasing very 
close to the hind part of the head, which is not definitely separate from the spinal 
region. 
Hence the parachordal region, or investing mass (iv.) is very short, only one-fifth 
the length of the entire skull; its moieties become the trabeculae (tr .) at the front 
third of the auditory capsules ( au .) The right and left bands unite below (Plate 10, 
fig. 3) but not above (figs. 1, 2); there they do project a little behind the capsules, 
but, below, this is hardly perceptible. Thus, the basi-occipital cartilage is incomplete. 
Moreover, it is arrested behind, for there are not only no condyles, but the cartilage 
stops in front of the place where they would be found. The sides, or ex-occipital region, 
and the roof or super-occipital, are arrested, entirely; thus there is no occipital ring. 
At this hind part the theca cranialis, or “dura mater” lies on an imperfect floor of 
cartilao’e, but for the rest of its extent the trabeculae lie rather outside than under it, 
and only meet and unite under the nasal capsule (na.) Thus the huge oval “ fontanelle ” 
reaches from end to end, and is only separated from the inferior or pituitary fontanelle 
by the thick lateral (trabecular) bands ; yet a secondary cartilaginous part forms a partial 
floor to the _pro-chordal part of the skull. The fore part of the skull has a peculiar ox¬ 
faced form, ending in two crescentic “ horns ; ” behind these horns it narrows gently, 
and then widens out in an even, rounded manner, the widening being due to the facial 
basket-work. 
The whole structure is, indeed, a generalised cranio-facial basket-work. There is no 
proper segmentation, but here and there the cartilage remains soft, and thus certain of 
the territories are marked out. 
That which is so puzzling in this little unossified skull is its primitiveness and 
simplicity ; for we are accustomed to cranial and facial structures that are differentiated 
from each other, and in which the facial arches are segmented into a number of parts ; 
these can be classified and named, a typical arch being made the measure of the rest. 
Here we have unenclosed land, and in such a generalised, common field, any balk, 
mound, stone, or bush may be useful as a landmark.* 
The “horns” are not the cornua trabeculae; these are suppressed in Myxine, but they 
belong to the palatine region, and so does at least half of the fore part of the cartilage 
| bounding the narrow cranial cavity. The two trabeculae are confluent in front, between 
* If the reader would follow the description and interpretation here attempted, it would be well for 
him to have the other papers of the writer before him, especially the following, namely :—“ Skull of 
Batrachia,” Part II., Phil. Trans., 1876, Plates 54-62; “Skull of Batrachia,” Part III., Phil. Trans., 
1881, Parti., Plates 1-44; and “Skull of Lepidostens Phil. Trans., 1882, Plates 30-38. The nomen¬ 
clature shall be as uniform as I can make it. 
