perforated at its base by the infraorbital foramen. The 
vertebral column was most peculiar. Ofthe seven cervical 
_vertebrze the first and sixth only were free, then came the 
“trivertebral bone” of Prof. Huxley, formed of the 
last cervical and first two dorsals; this was hinged 
ginglymoidly by the transverse processes on the next 
mass, which Was composed of the rest of the dorsal 
vertebra. The lumbar vertebrze were anchylosed to- 
gether, but not to the last (the 13th) dorsal. The centra 
of the vertebrae were only represented by a thin bony 
plate which helped to form the tube for the spinal cord. 
Prof. Huxley thinks that the joint in the dorsal region 
was connected with the respiratory process, Burmeister 
considers that it enabled the animal to retract its head. 
The pelvis was much as in the Armadillo ; the ilia were 
perpendicular. The symphysis pubis was slender, and is 
but infrequently preserved. The scapula possessed the 
characteristics of the class; the humerus had a supra- 
condyloid foramen, and the radius and ulna were not 
anchylosed, The ungual phalanges were all hoof-like. As 
in the armadillos and seals, the 4th metarcarpal bone 
articulated with the cuneiform as well as with the unci- 
form, and the 5th with the latter only; the pollex was 
absent. There was no third trochanter to the femur ; 
the astragalus was normal ; and there were four or five 
toes to the hind foot. 
Of the other extinct Edentata there are none closely 
allied to the Ant-eaters, Ant-bears, or Pangolins. In 
the Upper Miocene of Darmstadt an ungual phalanx has 
been found, which like those of Manis is longitudinally 
_ split ; it has been named Zacrotherium by Lartet. Teeth 
_ have been found since, much like those of Dasypus, so it 
could not have been a Manis. At Pikerme similar 
remains have been found, which have been named as 
parts of Ancylotherium. These animals must form a 
_ separate division of the Edentata. 
The Ungulata are the next great group to be consi- 
dered. In their teeth they all tend to the typical formula 
4m, a 44. The milk teeth are always 
a. Spe oT 
i. : c. > p.m. 
_ functional and generally remain until the animal is nearly 
fully grown. The limbs are formed for simple support 
_and progression, and there is no trace of clavicles. 
Except in the camels, the hoof encloses the ungual phalanx. 
Leaving Hyrax out of consideration it may be stated that 
the pollex and hallux are always absent. The class is 
found in all the world but the Australian region ; they 
have not been found fossil in strata below the tertiaries, 
_ but in the oldest of them. They are absolutely divisible 
into two sub-classes from characters indicated by, but not 
entirely dependent on the structure of the feet. This 
division was indicated by Cuvier, but developed by Owen 
and H.N. Turner, Jun. Owen introduced the terms 
_ Perissodactylata and Artiodactylata by which these sub- 
classes are best known. In the Perzssodactylata the 
middle toe is symmetrical, and there is typically one toe 
on each side of that, except in the Tapir and an early 
fossil Rhinoceros, Acerotherium. The astragalus has a 
single large anterior facet, entirely or mainly for articu- 
lation with the naviculare. There is a third trochanter to 
the femur, and there are never less than twenty dorso- 
lumbar vertebrz. The nasal bones expand posteriorly, 
and there is an osseous alisphenoid canal, as pointed out 
by Mr. Turner. In the Artiodacty/ata the axis of each of 
the feet is between the middle and fourth toe, and there is 
one toe outside each of these, but in the Peccary one is 
absent in the hind foot. The astragalus supports both 
the navicular and cuboid bones on nearly equal-sized 
facets. here is no third trochanter to the femur, and 
_ there are always nineteen dorso-lumbar vertebrae. There 
is no alisphenoid canal; the palate is completed by bone 
‘opposite the posterior molars. All these characters, espe- 
cially when taken in connection with the teeth, make it 
easy, from a few fragments of the skeleton, to identify the 
sub-class to which fossil members of the class belong. In 
the Perissodactylata the persistent premolars and the 
molars are very much alike, being all double, but in the 
Artiodactylata the premolars are single, and therefore they 
do not form a uniform series with the molars; there is 
a third lobe to the last lower molar, except in an antelope, 
Neotragus Saltiana, as lately proved by Sir V. Brooke. 
The earliest of the Perissodactylata are the Lophiodon- 
tide, from the lower Eocene. They are rather more 
generalised than the existing forms, as would be expected 
on the evolution hypothesis. The premolars are simpler 
than those behind. Coryfhodon, the oldest, is known by 
its teeth, of which it possessed the typical forty-four, and 
a femur with a third trochanter. Its molars had two 
ridges with conical apices, whence Owen gave it its name, 
from a specimen dredged off Harwich. The feet are not 
known unfortunately. 
Lophiodon itself is only known by its teeth and frag- 
ments of the skeleton ; the upper part of the skull has not 
yet been found. It is a genus of the early and middle 
Eocene only. Its dental formula was i. 3 (eS “pam, 3 m,> 
3 See 38 
The molars are representatives of the type which runs 
through the whole class. In the upper jaw each tooth 
presented an outer wall, which developed into two well- 
formed cusps ; running back from this were two trans- 
verse ridges, an anterior and a posterior, as they are 
termed. The anterior transverse ridge springs from in © 
front of the anterior cusp, the posterior from in front 
of the one behind ; these ridges, each by their curv- 
ing backwards, enclose a space named the anterior and 
posterior sinuses. The posterior transverse ridge is ab- 
sent in the premolars, The lower molars are simpler. 
The premolars were reduced in a manner which charac- 
terised the genus, the hinder part being cut away. 
Lophiodon is mainly found in the lacustrine deposits of 
the south of France ; there were several species, varying 
in size from a full-grown Indian rhinoceros to that of a 
hare. Leidy, from the deposits of Nebraska, found a 
tooth exactly resembling those of this genus; he has 
named it Z. occidentalis, but acknowledges the insufficiency 
of the evidence on which it is founded. 
The next animals to be considered were about the size 
of the hare. Pachynolophus differed from Lophiodon in 
having seven pre-molars and molars in the upper jaw 
instead of six ; the number in the lower jaw is not quite 
so certain, some having apparently six and others seven. 
The ridges of the teeth were less considerable and more 
broken into tubercles. In the London clay, near Herne 
Bay, a skull was found, named by Prof. Owen Hyra- 
cothertum, Mr. H. N. Turner was the first to point out 
that this animal was one of the Perissodactylata, and not 
as Prof. Owen at first supposed, allied to the Suidz, The 
teeth very closely resembled those of Pachynolophus, each 
transverse ridge developing intoa median smaller tubercle 
and a posterior larger one ; the pre-molars were also con- 
siderably smaller than those behind. From the resemblance 
of the teeth it is evident that the French genus Pachynolo- 
phus and the English Hyracotherium must be considered 
to be one. Péiolophus is the name of a genus given by 
Owen to a specimen obtained from the London clay off 
Harwich, together with a humerus, femur, and three meta- 
tarsals. The forty-four teeth were present. There is no 
reason to suppose that this genus is different from Hyra- 
cotherium, the shape of the teeth and the size being 
identical ; Prof. Owen himself states the possibility of their 
identity. The fact of there having been three metatarsals 
or metacarpals as they may be, found together, is a col- 
lateral one in favour of the animal having been Perisso- 
dactylate. From the above remarks it is evident that the 
names Pachynolophus and Pliolophus must be sunk in 
favour of Hyracotherium, as must also Lophiotherium, a 
name given by Gervais to an Upper Eocene specimen, 
known only by the mandible, 
