connexion. In Camptosaurus the supra-occipital bounded the foramen 
magnum above and in Edmontosaurus it is probable that it occupied the 
same position. 
The opisthotic portion of the exoccipital-opisthotic complex extends 
backward, with a slight inclination upward from the horizontal, and 
develops posteriorly a large forwardly hooked paroccipital process directed 
outward and downward beneath the squamosal process which it supports 
and resembles somewhat in general shape, and beyond which its pointed 
end projects. 
Above the lateral convexity of the occipital condyle is a concave 
surface in which are four foramina (Figures 26 and 27, IX, X, XI, jug. v, 
a. c. /., XII) to be referred to later, piercing the exoccipital. A strong, 
backwardly directed ridge, springing from the basi-occipital, bounds the 
convexity above and merges farther back into the lower marginal curve 
of the paroccipital process. Above the ridge the external surface is widely 
channelled longitudinally behind the fenestra ovalis (Figure 26, fen. ov.), 
and it is in this upper depression that the division between the opisthotic 
and the epiotic would be expected. In this channel was probably lodged 
the stapes whose anterior end closed the fenestra ovalis. 
Supra-occipital (Soc.). Figures, 4, 5, and 6. Extending back from the 
transversely angulated upper rim of the foramen magnum is an extensive 
flat surface of bone which lies between the paroccipital processes of the ex- 
occipital and broadens backward with an inclination upward from the 
horizontal. No indications of sutures are found in this region and evidently 
the supra-occipital which would be expected here in the midline, and the 
exoccipitals have completely codssified. It is impossible, therefore, to 
determine definitely the limits of the supra-occipital, but it is thought 
probable that the greater part if not all of this surface is supplied by the 
supra-occipital and that this bone entered into the formation of the 
foramen magnum above, confining the exoccipitals to the flanks of the 
opening somewhat as in Diclonius mirahilis as figured by Cope (1883, 
pi. VII). 
A definite ridge extends backward from the foramen magnum along 
the midline of this supposed supra-occipital surface, and anteriorly on 
either side of the ridge the bone is widely excavated upward behind the 
foramen magnum, the exoccipitals flanking the excavations externally. 
The back border of this surface is straight transversely for some 
distance outward from the midline and then suddenly acquires a greater 
backward protrusion which curves evenly outward into the paroccipital 
process. On the upper face of the protruded border on each side are 
curved impressed lines (Figure 8, c) which may indicate the postero- 
external limit of the supra-occipital. 
The anterior contact of the supra-occipital with the parietal is not 
seen, but posteriorly there is a space between these two bones, a low space 
which extends outward for a short distance, about 20 mm., from the 
midline beneath the squamosals. 
Viewing the skull from below, its extension backward from the foramen 
magnum is conspicuous. A noteworthy feature, in a posterior aspect, is 
the lowness of the parieto-squamosal arch in comparison with its breadth, 
as well as the smallness of the parietal contribution to the arch. Beneath 
this arch, following its curve, and closely applied to it, except near the 
