FROM THE CAPE OF GOOD HOPE. 197 
In Ichthyosaurus the pulp-cavity may be partially obliterated in 
the root while it is retained in the basal part of the crown of a 
tooth*; but the base is sapped by the successional tooth, which 
penetrates its predecessor and seems to push it out?. The root, in 
fully formed or old teeth, contracts slightly at the implanted end. 
In Iguanodon the tooth-root is less broad than the crown, and 
assumes a subcylindrical shape; but such root seldom contracts 
as it descends, and never dwindles to a point; it is invaded and 
pressed upon by the crown of the successional teeth, developed in 
germ-sockets at the intervals left by the narrow roots of the teeth in 
place, which roots undergo absorption as the tooth they support is 
in course of being displaced by its successor §. 
The Pliosaur and Iguanodon exemplified, in my experience, the 
extremes of deviation from the developmental phenomena of the 
teeth of existing Crocodiles and Lizards prior to the discovery of the 
Theriodont type of Reptilia. 
In the few rare, and at that time unique, examples of the order 
described in my ‘Catalogue of the Fossil Reptilia of South Africa,’ 
but one instance showed, by its fractured state, the implanted part 
of the canine || ; and this did not appear to afford adequate ground 
for adding the condition of tooth-root and limited succession of teeth 
to the other mammalian characters assigned to the order]. But I 
now know that a long deeply implanted root of the canine tapering 
to a blunt point, with the widest part of the pulp-cavity at the base, in 
or near to the crown, is a natural structure and an ordinal charac- 
teristic by which the Theriodontia differ from the Dinosauria and 
Enaliosauria, and in a. still greater degree from the Crocodilia; and 
this character must be added to the serrate and trenchant border of 
the crown of the canine, to the entocondylar perforation of the 
humerus, and to the general carnivorous type of the dentition—in 
which the incisors, defined by position, are more equable in size than 
in Crocodilia, while the canine, large and laniariform, is single on 
each side of both jaws, with the crown of the lower canine crossing 
in front of the upper one; this, moreover, is followed by a series of 
small, subequal, subtrenchant, pointed molars. 
In all such groups, as our knowledge thereof extended, we com- 
monly became acquainted with and perplexed by aberrant forms ; 
and such were diminutive species associated under the generic name 
of Procolophon, and which, in my ‘ Catalogue ’ of 1876, I left at the 
faz-end of the type species of Theriodontia, while at the same 
time pointing out in Procolophon minor, for example, the “linear 
osseous narial septum” (p. 26), and duly representing that differen- 
tial character from the Mononarials in pl. xx. fig. 11—noting also 
* Odontography, p. 376, pl. 73. fig. 8 @. 
+ Ib. p. 280, pl. 73. fig.77: and see ‘Monograph of the Fossil Reptilia of the 
Cretaceous Formations,’ Pal. vol. 1851, pl. iv. figs. 7, 8, 9. 
t Jb. p. 249, pl. 70. figs. 1, 2, 4. 
§ ‘Monograph on Wealden Reptiles.’ Palxontographical volume issued in 
1874, p. 3, pl. i. fig. 8, a, b, 0. 
|| Lycosaurus pardalis, op. cit. p. 16, pl. xiv. fig. 1, ¢. 
§| Op. cit. pp. 15 & 75. 
Q.J.G.8. No, 138. P 
