3998 
faces of the maxillary and premaxillary bones, I said that: “In se- 
lachians the anterior margin of the nasal velum of GEGENBAUR’S 
descriptions would seem to represent the breathing-valve of teleosts.’’ 
This is evidently an error, if my present conclusions are correct. In 
that work I further said that: “van WisH8 says of these maxillary 
breathing-valve bones of Polypterus that they are functionally dermo- 
palatines’’; which is not correct. What van WısHE says is: ‘‘Dieser 
Teil fungirt also als Dermopalatinum’’; and the “‘part’’ referred to 
is the anterior portion of the ectopterygoid and not the so-called 
vomer bone. 
The orbital process of Chiamydoselachus is a large process rising 
from the dorso-mesial edge of the palato-quadrate at about the 
anterior third of its length. It is directed dorso-mesially and is not 
continued ventrally, as a ridge, onto or across the ventro-mesial 
surface of the cartilage, as it is in both Hexanchus and Heptanchus 
(GEGENBAUR), this apparently being due simply to a large dental 
sroove which, in Chlamydoselachus, here cuts across the base of the 
orbital process and gives to. it a sharp ventro-internal edge. The 
process is everywhere covered with tough connective tissue, excepting 
only on that portion of its mesial surface that articulates with the 
neurocranium, this latter surface being covered with a layer of peri- 
chondrium, as GEGENBAUR says that it always is in other plagiostomes. 
The process fits into a sac-like cap of connective tissue, as in Hept- 
anchus (GEGENBAUR), and this sac, which is the “pad of capsular 
tissue’ of Goopry’s descriptions, is strongly attached to the anterior 
wall of the orbit, and also, but much less strongly, to the palato- 
quadrate. The process is not, itself, directly attached by ligament 
to the orbital wall, as GarMAN describes and figures it, my obser- 
vations thus confirming GoopHy’s statement to that effect. 
The process articulates with a large facet in the orbit, the process 
being so large that it occupies, in the limits of its action, the deeper por- 
tion of the orbit from its anterior wall back to the base of the eyestalk, 
and from its ventral edge up to the overhanging roof. ‘The process 
forces the nervus opticus forward against the anterior wall of the 
orbit, where a groove is found for its protection, and forces the eyeball 
away from the mesial wall of the orbit; and as the antorbital and 
postorbital processes of the chondrocranium are elongated to give 
support and protection to the anterior and posterior surfaces, respec- 
tively, of the eyeball, the well known deep orbit arises. ‘The process 
