369 
to Fucus, having still earlier given the name “l’espace postfacial du 
eräne” to what is apparently the homologue of the entire chamber of my 
descriptions. WınsLow also, in 1898, described this same chamber in 
young larvae of Amblystoma. This all entirely escaped my notice at 
the time that I was engaged on my latest published work (1909) relating 
to this subject, for I then left almost entirely out of consideration 
all vertebrates above amphibians, excepting only Lacerta and man. 
My work on fishes, taken in connection with Gaupp’s and Fucus’ 
work on higher vertebrates, thus leads me to conclude that the selachian 
eyestalk presents certain features which warrant the assumption 
that it is a mandibular or premandibular branchial ray, and that it is 
the homologue not only of the ascending process of the amphibian 
palato-quadrate and of the reptilian antipterygoid, but also of the 
mammalian ala temporalis. This element, thus homologized, of the 
vertebrate skull is a very variable one. It is attached by one end to 
the neurocranium in most selachians, but it is not so attached in 
Zyaena, and it is said by Harman to be absent in Seyllium. It is never, 
in selachians, attached to the palato-quadrate, so far as I can find. 
In Lepidosteus it is represented entirely by membrane, and in such 
teleosts as I have examined it is usually represented largely or wholly 
by membrane but may be wholly represented by bone. In amphibians 
it is attached both to the palato-quadrate and to the neurocranium. 
In the chondrocranium of Lacerta it is attached by its lower end to a 
detached portion of the palato-quadrate, while in the chondrocranium 
of higher vertebrates it is usually attached by its lower end to the 
cranial wall, but may be, at certain stages, wholly independent of 
that wall. It is always related, in all these vertebrates, either to a 
processus basalis, to a processus basipterygoideus, or to a part of the 
cranial wall that must correspond to the latter process; and this 
latter process would seem to have been developed, independently, 
in relation to the processus basalis and not from the suborbital shelf 
of certain selachians, as Verr (1911) has suggested and as Gaupp 
accepts as probable (1910, p. 338). And these two structures, the 
antipterygoid and the basipterygoid process and their several 
homologues or equivalents, always form a portion of the external 
bounding wall of what is either an extracranial space or an intradural 
cavity which lodges the whole or a part of the trigemino-facialis gang- 
lionic complex; and this space or cavity may become secondarily 
included in the cranial cavity proper of the prepared skull. 
Anat. Anz. Bd. 45. Aufsätze. 24 
