374 MR. ST. GEORGE MIVART ON THE VERTEBRATE SKELETON. 
9. What is the essential nature of ribs, transverse processes, and sternum ? 
Dr. Cleland, in the paper before referred to, has, I think, thrown much light upon this 
question; and with many of his arguments and determinations I fully agree. I do so 
especially as to the subordinate importance, in questions of serial or general homology, 
of the mode of ossification of parts or of their points of attachment to the vertebral 
column. With him, I believe that “ribs” are parts of the skeleton the primary office 
of which is to support the outer wall of the trunk, and I think the degree of segmenta- 
tion of such structures is of very little consequence morphologically. 
As to the significance, however, which is to be attached to the fact of parts being or 
not being “in series” I cannot at all agree with him; for I am at a loss to know what 
better or other guide we have in matters of serial homology. 
All will admit that that part of the rib which intervenes between its head and tubercle 
in one animal is the general homologue of the part which so intervenes in another ani- 
mal; and yet the researches of Rathke and Professor Huxley seem to show that some- 
times it isa primary part of the rib, sometimes a mere outgrowth. Again, all will admit 
that the rudimentary neural arches of the tail of the dog are the serial homologues of 
the fully developed ones of the trunk; yet, as Dr. Cleland himself tells us*, the former 
are merely exogenous structures. "These, and many other instances it would be easy to 
bring forward, show that diversity as to mode of development is no bar to the legitimacy 
of serial or general homologies. 
Just as little can relations of functions claim a preference as a mode of determining 
such homologies. Otherwise the neural rudiments of the os coccygis could not claim to 
be serial representatives of those of the trunk, or caudal transverse processes to represent 
parts of the viscera-encircling girdles in front of them. Thus the fact of parts being 
“in series" is most important in determining serial homologies, and, when carefully 
used and checked by comparing results obtained in many different forms, is the best 
available guide. 
I believe that all transverse processes, ribs, and sternal structures are elements of one 
system of parts, the primary office of which is to surround the visceral cavity (fig. 2, P, 
t, tr, tr’, e, er’, vr, sr, & s), that they are essentially parts belonging to the external 
lamina of the ventral plate of the embryo, and that their serial relationships, as whole 
arches and as component parts of such arches, are revealed by the fact of serial position 
in such arches or parts of arches. 
As to the conditions presented by the hard parts forming these arches, we find com- 
monly in the highest animals a rib with a head and tubercle, the latter articulating with 
a transverse process. : 
In many of the Amphibia we find a rib bifurcating proximally and more or less distally 
also, and articulating, by the branches of its proximal bifurcation, with two superimposed 
transverse processes (Plate LIII. figs. 14 & 15, £, c). 
In many fishes we find two series of ribs, one placed above the other; and sometimes 
these are attached to two more or less distinct, superimposed, transverse processes (Plate — 
= * Loc. cit. p. 120. 
