658 
ME. W. K. PAEKEE ON THE STEUCTTTEE AND 
invests the lower face of the cartilage both in the pterygoid and metapterygoid regions. 
This triacanthous plate binds on to the inner face of the quadrate (q.), and the under face 
of the pedicle (pd.), and of the lateral pterygoid bar of cartilage. 
Then, except where it is invested by the maxillary ( mx .), which runs to the inner side, 
almost meeting the pterygoid, the three regions of the palatine cartilage seem free from 
ectostosis. With great care, however, there may be found a style attached to the hinder 
face of the ethmo-palatine below ; it runs outwards and forwards, and sets a pedate base 
against the antero-external angle of the parasphenoid ( pa.s .) : this is the ectosteal pala- 
tine (figs. 3 & 7 ,pa.); it exists in the Batrachia generally, but not in Dactylethra 
(Plate 59. fig. 2)* 
The above, with the pair of bones in each columella, and the “ articulare ” in each 
mandible, are all the deep-layer ossifications I can find in this stage ; they are bones 
that can be easily identified with their homologues in Osseous Fishes, Sauropsida, and 
Mammalia. 
The “ investing bones ” are of great interest in this stage and in this type ; the largest 
and earliest of these is the parasphenoid (fig. 3, pa.s.). This huge bony shield is not 
like what we see in the Batrachia generally, but resembles that of an adult “ Perenni- 
branch,” or a young “ Caducibranch ” Urodele. Its hinder half is the narrower ; it 
gradually loses breadth behind, and then its rounded hinder margin sends backwards a 
sharp spike. In front it is also spiked, and then, for a short distance, is only one third 
its full breadth, but, by a sudden step on each side, attains it rapidly. Its narrowing 
takes place suddenly again, where the broad foot of the “ pedicle ” joins the trabecula. 
Beneath its front spike there is no vomer. A somewhat larger space is covered by the 
roof-bones above (fig. 2). This large bony part is very similar to that below ; it is, 
however, convex along two lines and somewhat depressed along the middle, whereas the 
parasphenoid is convex generally. The upper tract is spiked in front, but from a bor- 
rowed source, the superethmoidal. It is winged in front ; but the upper belong to it, 
whilst below they are borrowed. Behind, the roof is crenate and not spiked. This part 
is also formed, not from one, but from four centres, namely, the frontals, which are twice 
the length and breadth of the others, the parietals ( f.p .). The anterior frontal wing 
(prefrontal process) binds the ethmo-palatine above, just as the palatine style binds it 
below. As the bony roof is larger than the floor, so the space to be covered (Plate 60. 
fig. 5) is also larger. As in Dactyletlira , there are no secondary fontanelles to be 
covered by the parietals, it is all one pyriform space. 
The narrow walls and spurs between and around the nasal mucous membrane, expos- 
ing it largely in the chondrocranium (Plate 60. figs. 5 & 6), are largely supplemented by 
investing bones. The main pair are the nasals (Plate 61. figs. 2 & 4, n.) ; they are large 
bones, nearly semicircular, but their obliquely placed transverse margin is deeply notched 
* Professor Huxley missed this hone in the adult (art. “ Amphibia,” p. 756). The unlooked-for “ wings ” 
to this part in the huge parasphenoid of the adult (Plate 62. fig. 2, pa., pa.s.) struck me as abnormal, and led me 
to make inquisition in the young. 
