1847.] OWEN ON EXTINCT ANTHRACOTHERIOID QUADRUPEDS, 133 
the Proboscidian as well as of the Perissodactyle groups. In addition 
to the osteological characters derived from the skeleton of the feet, 
by which Cuvier illustrated the affinities linking together the Perisso- 
dactyle pachyderms, he first pointed out a common character in their 
femur, the presence viz. of the third trochanter. This character is 
absent not only in the ‘pachydermes 4 doigts pairs’ of Cuvier, but 
also in the Ruminants. In addition to the common characters of the 
digestive canal, viz. complexity of stomach and simplicity of czecum, 
which have seemed to me to justify the association of the even-toed 
Bellue of Linnzeus with his Pecora, there is a positive osteological 
one which has not received the attention it merits, in the question of 
the affinity of the ‘pachydermes a doigts pairs’ to the Ruminantia 
of Cuvier. Both these groups, or, in other words, the Artiodactyle 
Ungulates, have the same number of what the anthropotomist calls 
‘true vertebree,’ 2. e. moveable or unanchylosed vertebrze, intervening 
between the skull and sacrum. The vertebree of the cervical region 
are of course seven in the Artiodactyles, as in other Ungulata and 
in most Mammalia: hut the rest, viz. the dorsal and lumbar vertebree, 
are nineteen in number in all the Artiodactyles, neither more nor less ; 
whilst they exceed that number considerably in all the Perissodactyle 
group, where they range between twenty-two (Rhinoceros) and twenty- 
nine (Hyrax). What seems to have masked both the fact and the im- 
port of the constancy in the number of the dorso-lumbarvertebral series, 
as evidence of the natural character of the Artiodactyla, is the variable 
number of the ribs in some of the species, which e. g. are reckoned 
at fifteen pairs in the Hippopotamus and twelve pairs in the Camel. 
And the value of this distinction has been exaggerated, through the 
prevalence of the anthropotomical conception of the ribs as indepen- 
dent bones, quite distinct from the vertebree, and not as equivalents 
with the neurapophyses or other autogenous but anchylosed vertebral 
elements. The discovery of rudimental ribs (pleurapophyses) at- 
tached to the ends of the lumbar diapophyses in the foetal Pig and 
some other quadrupeds*, which afterwards become anchylosed, and 
the known pleurapophysial nature of a part of the so-called perforated 
transverse process of the cervical vertebrz, show how arbitrary and 
essentially incorrect is that definition of a dorsal vertebra which calls 
it one that supports ribs. It is convenient, no doubt, in comparative 
tables of vertebree, to give the number of such vertebree of the trunk 
as habitually support their pleurapophyses without anchylosis; but 
_the differences sometimes occurring in this respect within the limits of 
the same species must cease to have their importance over-estimated 
when the true nature of a rib is recognized. Mr. Eyton’s interesting 
observation of fifteen dorsal and four lumbar vertebre in a Chinese 
Boar, and of thirteen dorsal and six lumbar vertebree in an African 
Sow+, for example, depends on the varying extent to which the lum- 
bar pleurapophyses have become anchylosed in those varieties. The 
actual number of dorso-lumbar vertebrze does not vary ; it is the nor- 
mal one in the Artiodactyle order, viz. 19. So also if we examine the 
* J. Miller, Anatomie der Myxinoiden, 1834, p. 238. 
t+ Proceedings of the Zoological Society, 1837. 
