242 
cavity in my 80 mm. specimen has become cartilage in the 149 mm. 
specimen, and as this roof ends anteriorly with a free edge, the cavity 
opens, its full height, into that part of the cranial cavity that lies 
immediately anterior to it. In the adult, the anterior edge of the 
roof of the cavity bends downward and fuses, excepting in its median 
portion, with the cartilage of the basis cranil, the large recess of the 
149 mm. specimen thus becoming the nearly closed cross-canal of 
SAGEMEHL’s (1883) and my own descriptions of the adult. The 
saccus vasculosus, in the adult, still projects into this cross-canal, 
but the pituitary vein has been largely pulled out of it, this appar- 
ently being due to a shifting in the position of the pituitary foramen, 
which foramen, for some reason that is not evident, lies, in embryos 
(Vert, Fig. 4, Venenloch), considerably anterior to the cross-canal, 
not far from the foramen opticum. ‘The lateral ends of the cross- 
canal are accordingly closed, and each end is separated from the 
corresponding trigemino-facialis chamber by only a thin layer of 
bone. SAGEMEHL homologized this cross-canal of Lepidosteus with 
the canalis transversus (pituitary canal) of selachians, a conclusion 
which Veıt accepts as probable but which I considered as not wholly 
correct because of the course of the pituitary vein. 
Comparing these conditions in Lepidosteus with those in 
Chlamydoselachus it is evident that, if the trabeculae of Chlamydo- 
selachus had been laid down in the line prolonged of the parachordals, 
the tip of the notochord would have retained its actual position 
immediately posterior to the pituitary vein, and hence at the hind 
end of the tough membrane that covers the pituitary canal. Two. 
suppositions can then be made regarding the development of a roof to 
the pituitary canal: first, that chondrification of the tough membrane 
that covers the canal in Chlamydoselachus gave rise to the roof of the 
cavum sacci vasculosi of Lepidosteus, and that the saccus secondarily 
acquired, by indentation, a position ventral to that roof; and second, 
that the roof was formed by the chondrification of portions of the 
loose connective tissues that, in Chlamydoselachus, overlie the tough 
membrane. And this latter supposition seems much the more prob- 
able, for, in Acanthias, as above described, the tissues that, in 
Chlamydoselachus, surround the pituitary vein have apparently been 
completely chondrified, and yet there is a long cartilaginous shelf 
which overhangs the hind end of the pituitary fossa much as the roof 
of the cavum sacci vasculosi overhangs that cavity in Lepidosteus. 
