64 
THE VOYAGE OF H.M.S. CHALLENGER. 
of bone quite separate from the peduncle, transverse process, and superior articular 
process ; but the two laminae, not united mesially and posteriorly into a single spine, were 
separated from each other by a mesial cleft (fig. 2). In both the Bushman and the male 
Esquimaux the laminae had united posteriorly into a mesial spine, but they and the 
inferior articular processes were not fused with the pedicles, so that the posterior part 
of the neural arch formed a separate piece of bone. The very imperfect spine of a 
Sandwich Islander, recently presented by Dr. G. W. Parker of Waiahia, Oahu, which is 
not included in the preceding thirty skeletons, exhibited a defect in the neural arch of 
the 5th lumbar, similar to that seen in the Bushman and Esquimaux. In one of the 
Negro skeletons the right half of the neural arch was normal, but the left lamina, with 
its inferior articular process, had remained distinct and separate from the spine, as well as 
from the pedicle of its own side. These five specimens show that the laminar part of 
the neural arch had arisen in them from a pair of ossific centres, quite distinct from the 
centres from which the pedicles took their origin ; and that whilst the inferior articular 
processes belonged to the laminar part, the superior belonged to the peduncular part of 
the vertebra. As this condition was only seen in the 5 th lumbar vertebra the question 
naturally arises, does the neural arch of this bone on each side normally develop 
from two ossific centres, laminar and pedicular, which subsequently fuse with each other, 
though in these five skeletons they had remained distinct ; whilst in the other four 
lumbar vertebrae only a single pair of ossific centres are formed for the production 
of both the peduncular and laminar part of the arch. It is known, for instance, that the 
anterior division of the transverse process (parapophysis) in the 7th cervical vertebra 
is developed from a distinct centre, which may sometimes acquire the form and dimensions 
of a cervical rib, whilst the parapophysis in each of the other six cervical vertebrae arises 
by an extension into its cartilage of ossific material derived from the primary centres of 
ossification of the bone, so that this process in the cervical series does not develop 
uniformly in all the vertebrae. It is possible that the neural arch of the 5th lumbar 
vertebra may normally develop on a different plan from that of the other lumbars, so as 
to permit of such a separation of the laminar part of the arch as was found in these 
specimens. The development of the neural arch in this bone is worthy, therefore, of 
special investigation by the embryologist. 
In the General Summary of the characters of the aboriginal Crania described in the 
First Part of this Report, I pointed out (p. 118) that a larger proportion of important 
variations from the usually described arrangements occurred in them than would have 
been found in a corresponding number of the skulls of the white races. With equal 
truth a similar remark may be made on the variations in the Spinal Column which I have 
just described. I am not able to state, numerically, the proportion in the white races 
in which variations in the form of the spinous processes of the cervical vertebrae, the 
occurrence of a supernumerary vertebra in the dorsi-lumbar region, or the development 
