No. 2.] THE HARD PARTS OF THE MAMMALIA. 215 
animals. In the first place, the effect is not due to the meta- 
pophysis, but to the anterior extension of their bases. This is 
demonstrated by the fact that the articulations are equally 
developed in the anteater and in the armadillo, although in the 
former animal the metapophyses have not the enormous length 
that they have in the latter. That they are due to anterior 
extension of the base is shown by the fact that in various mam- 
mals where they are developed as largely as in the anteater (e.g. 
the mole), but where they have the usual position, the additional 
vertebral articulations do not exist. In TZalpa europea the 
metapophysis overlaps the neural spine of the vertebra in front 
of it very little. In both Myrmecophaga bivittata and Priodontes 
maximus its anterior extremity is opposite the anterior border 
of the neural spine of the vertebra in front. The anterior trans- 
ference of the metapophyses could then only have been accom- 
plished by longitudinal muscular strain. This would be pro- 
duced by strong upward curvature of the column, especially of 
its lumbar region, and strain on the intervertebral ligaments and 
muscles. The zygosphenal arched plate would be thus produced, 
its effect being to protect the spinal cord, while the neural arches 
were strained apart. The metapophysis, as an important mus- 
cular insertion, would be subjected to an anterior drawing strain, 
which would cause its base to advance with the advance forwards 
of the neural arch, and finally to notch, and at last to divide the 
arch in front of it, by pressure caused by its apex on the straight- 
ening of the column, thus cutting off the episphen. This action 
would be distinctly aided by the increasing width of the meta- 
pophysis at the base, which spreads the zygapophysial faces out- 
ward and produces a longitudinal angle in the lateral parts of 
the surfaces in contact, cutting off the zygosphenal from the 
zygantral surfaces. 
We find that just such flexure of the vertebral column as is 
above inferred, is a characteristic habit of most Edentata, espe- 
cially of the anteaters, armadillos, and pangolins (Manidz). 
They roll themselves into a ball-like form, and retain the posi- 
tion against attack with great energy. But here we are con- 
fronted by the fact that the Manidz possess this habit, and yet 
totally lack the complex vertebral articulations of the American 
Edentata (Xenarthri). If we refer the energetic flexures of the 
vertebral column to fossorial habits, we are met by the same 
