44 THE VOYAGE OF H.M.S. CHALLENGER. 



bone. The spine of the atlas was rudimentary ; in the axis it was massive ; in the 

 3rd cervical it was short and it gradually increased in length to the 7th cervical. 

 A strong process, distinct from the transverse process, projected backwards for nearly 

 half an inch from the pedicle of the 7th cervical immediately below the articular process, 

 and indications of a simdar process were seen in both the 6th and 5th. The bodies of 

 the cervicals were mesially keeled on the ventral surface, and in the 7th the keel was 

 elongated into a plate-like hypapophysis, which was very projecting in the adult Messier 

 Channel male. 



The dorsal vertebra} had relatively short transverse processes, which were the 

 longest in the more anterior dorsals, but diminished in length in the posterior dorsals 

 so as to be scarcely recognisable in the 14th and 15th. Anapophyses were present 

 in all the dorsals ; in the more posterior they were elongated and styloid, in the middle 

 and more anterior they were stunted and tubercular, and in series with the strong back- 

 ward-projecting process from the pedicle of the 7th cervical. Metapophyses projecting 

 forwards from the anterior articular processes were especially seen in the middle and pos- 

 terior dorsals. The spines projected backwards ; they were elongated and strong in the 

 anterior vertebra?, but gradually diminished in size from before backwards. The bodies of 

 the more anterior and more posterior dorsal vertebrae were keeled on the ventral surface. 



The lumbar vertebrae had transverse processes which projected downwards, outwards, 

 and forwards; in the 1st lumbar they were very short, but they increased in length 

 to the 4th. The spines were short and hatchet-shaped. Anapophyses had almost 

 entirely disappeared ; metapophyses were present in all of these vertebras. The ventral 

 surface of the bodies was keeled. 



Sacrum. — In the adult male from the Messier Channel the 1st and 2nd sacral 

 vertebrae were completely fused together, the lamina? of the 2nd were fused with those 

 of the 3rd, but the spines were not, and between their bodies the intervertebral disc was 

 partially present. The 3rd sacral vertebra was fused with the 1st caudal by the pair 

 of processes situated below the articular processes. The greatest breadth of the base 

 of the 1st sacral vertebra was 48 mm. and the length of its body was 32 mm. The 

 2nd and 3rd were smaller bones, especially in breadth ; they were all fully ossified 

 In Arctocephalus gazella fusion of the sacral vertebrae had not taken place. 



The caudal vertebra? diminished in size from the 1st to the 11th. The 1st, 2nd, 

 and 3rd possessed laminae and spines, the 4th had a neural groove, but the rest consisted 

 only of the bodies, and the two terminal segments of the tail were only 6 and 4 mm. 

 long respectively. 



Ribs. — There were fifteen pairs of ribs, of which nine articulated with the sides of 

 the sternum. The head of the 1st rib articulated only with the body of the 1st dorsal 

 vertebra, but from the 2nd rib backwards to the 12th, both inclusive, the head articulated 

 with the bodies of two vertebrae. The three most posterior ribs again articulated, each 



