214 COPE. [Vou. III. 
side, opposite to a corresponding facet on the lower side of the 
episphen. These two articulate with corresponding facets of 
an anterior prolongation of the metapophysis of the vertebra 
next posterior! In the Priodontes maximus the transition be- 
tween the normal articulations and those just described may be 
observed as we pass from the anterior dorsal region posteriorly. 
First the roof of the neural arch projects anteriorly beneath 
the arch of the preceding vertebra, forming a zygosphen of a 
very depressed type, combined with the prezygapophyses. The 
postzygapophysial surfaces are also continuous with the inferior 
side of the neural arch, or zygantral surface. The zygapophysial 
surfaces soon become distinct, but are not supported by distinct 
processes anterior to the last lumbar and first sacral vertebrae. 
The metapophyses begin to develop well anteriorly on the dorsal 
series, but they do not articulate with the zygantrapophysis in 
front of them until the sixth dorsal. On the seventh dorsal it 
joins a single large facet above and external to the postzyga- 
pophysis. On the eighth this facet, z.e. the roof of the neural 
canal, is deeply fissured by an uplooking ridge which separates 
the prezygapophysis from the zygosphen, and the episphen has 
its origin. On the ninth vertebra this ridge widens, and on the 
eleventh and twelfth it becomes distinctly part of the base of 
the metapophysis, which articulates above with the episphen, 
and below with the zygantrapophysis. 
This succession of development of these articulations proba- 
bly indicates the manner of their orign. Both the episphenal 
and zygantrapophysial articulations are due to the splitting of 
the roof of the neural arch on each side behind, by a keel of a 
zygosphenal plate which underruns the usual roof from the 
vertebra behind. It is a significant fact that this splitting, ze. 
the development of the angular ridge on each side the zygo- 
sphenal roof, first appears in both the anteater and armadillo, 
on the first vertebra which is disconnected with the sternum, or 
without hemapophysis ; and which is therefore susceptible of 
the greatest amount of vertical movement. The prolongation 
of this ridge is really the prolongation of the base of the meta- 
pophysis. 
We can now, I believe, associate this peculiar vertebral struc- 
ture as an effect with anteroposterior strains as a cause, in these 
1 Flower, Osteology of the Mammalia, 1885, p. 62. 
