244 
notochord here remained on the dorsal surface of the cartilage of the 
basis cranii, as it is also shown in PArker’s (1882) figures of trans- 
verse sections of a 21/, inch (63 mm) specimen. The hgamentum 
longitudinale dorsale inferius must accordingly here, if it persists, 
lie free in the cranial cavity and doubtless be capable of the ossi- 
fication without previous chondrification that here takes place in 
certain teleosts. In the posterior portion of this postpituitary portion 
of the cranial cavity, Veit shows, in his figure 4, what is said to be 
a space that lies between dorsal and ventral floors of this part of the 
cranial cavity, the ventral floor being said to be formed by the para- 
chordals and the dorsal floor by horizontal processes of the Pleuro- 
oceipitals (exoccipitals) which project mesially and meet in the median 
line. The space between these two floors is definitely said to open 
to the exterior on either side of the chondrocranium, thus traversing 
the basis cranii from one side to the other, and the descriptions and 
the figure would seem to equally definitely show that it also opens, 
anteriorly, directly into the cranial cavity, thus becoming a recess 
of that cavity. The two floors of the cranial cavity here referred to 
by Vurr are shown by PARKER in his figures of transverse sections of 
the 21/, inch specimen above referred to, but the space between the 
two floors, as there shown, extends, on either side, only to the noto- 
chord, and not from one side to the other of the chondrocranium. 
In my 50 mm. specimen I find conditions strictly similar to those 
shown by PARKER, but in my 80 mm. specimen the notochord is, in 
places, partly broken down, and the space between the two floors 
there has the appearance of extending from one side to the other of 
the chondrocranium, as Vert describes it. In neither of my specimens 
does the space open anteriorly into the cranial cavity, and in two 
adult specimens that I have also examined there is also no slightest 
indication of a recess in the cranial cavity such as is apparently shown 
by Verr in his figure. My sections accordingly show that the two 
floors of Vurr’s descriptions are simply the dorsal and ventral para- 
chordal bands of that author’s later (1911) descriptions of younger 
specimens of Lepidosteus; that the space between these two floors 
extends normally only to the notochord, but may perhaps become, 
at certain stages, by the breaking down of the chordal tissues before 
the formation of the basioccipital bone, a transverse space extending 
from one side to the other of the neurocranium; and that this space, 
at least in living specimens, probably at no time communicates with 
