SIUFELDT.] OSTEOLOGY OF THE EREMOPHILA. 637 
atloid articulation so thoroughly re-enforced by the thick muscles that 
surround it. 
A square bony plate projects from below in the atlas, more anterior 
than any other part of the bone, that covers the atlo-axoid articulation 
in front. 
The arch that connects the neurapophyses is broad and smooth, and 
assists greatly in the protection of the myelon between the two. bones. 
The odontoid process on the awis is concave in front, flat behind, with 
a roundishsummit. It averages one millimetre in len eth, and is directed 
slightly backwards. The articular surface at its base is reniform in 
outline, the centrum that supports it being contracted below. The 
postzygapophy ses show faint traces of anapophysial tubercles; these 
are better marked in the latter cervicals. The last or thirteenth verte- 
bra has freely suspended from beneath each diapophysial articular sur- 
face a rudimentary pleurapophysis that averages about two millimetres 
‘inlength. These little bones represent the only true cervical ribs, though 
we must admit here that in several individuals we found the first pair 
of dorsal pleurapophyses unconnected with the sternum by the usual 
hemapophyses, and ending in pointed extremities. Should such a speci- 
men alone be examined, we would have to recognize fourteen cervical 
vertebre, the last two bearing free pleurapophyses, but the common rule 
must dictate here as elsewhere, and the condition just mentioned be 
reckoned as the exception. 
Dorsal vertebra, vertebral and sternal ribs, sternum—(P1. IV, Figs. 22, 
24, 27, and 38).—The number of vertebrze devoted to the dorsal portion 
of the spinal column in Hremophila seems to be invariably five. They 
are easily detached one from another, and after ordinary maceration of 
the skeleton drop apart almost as readily as the cervical vertebree, so 
that during life there is at least quite a little amount of free movement 
among these bones. 
The neural canal, as it passes through this series, starts with the 
transverse ellipse as we left it in the last cervical, in the vicinity of the 
dorsal expansion of the myelon, to terminate nearly circular, and much 
diminished in calibre, in the ultimate segment of the sacral extremity. 
The neural spines form. by their interlocking a continuous ridge 
above. The thickened crest of this ridge is produced by what we will 
call the arrow-head joint, a true schindylesial articulation to be found in 
many of the class. The superior margin of each spine becomes pointed 
anteriorly, extends forward, and is received into a fissure of the poste- 
tiorly produced superior margin of the neural spine of the vertebra next 
beyond it. This arrangement has the appearance of so many little arrow- 
heads placed in similar juxtaposition, and constitutes one of the elements 
of stability of the dorsal vertebre in this bird. The open spaces remain- 
ing among the bodies of the spines below, between their produced crests 
and the several neural arches, are filled up by connecting ligament and 
membrane. 
The diapophyses of the dorsals are a very much horizontally flattened 
series. They are all slightly tilted upwards, the anterior ones being the 
broadest and shortest, and the ultimate one, by a gradual departure in 
this regard from the first, the narrowest and longest. In the middie of 
the series, moderately well developed and antero-posteriorly produced 
metapophysial ridges are found limiting the diapophyses externally; 
they do not reach from one vertebra to another. The pneumatic for- 
amina at the bases of these processes are very minute and scarcely dis- 
cernible by the naked eye. 
The inferior diapophysial facettes for the pleurapophysial tubercula 
