22 PEOCEEDINGS OF THE NATIONAL MUSEUM vol.73 



In the Phoca the spines, especially caudad, are but little more pro- 

 nounced than the other processes. The change in the direction of the 

 articular surfaces of the zygapophyses between the eleventh and 

 twelfth thoracics as mentioned for Zalophus is much more abrupt 

 and especially caudad and in the lumbar region the zygapophyses- 

 of each vertebra are situated much farther apart — relative to size, 

 over three times farther. The interlocking is such as not to prevent 

 as marked concavity in the outline of the dorsum as the limitations 

 of the sternum will permit, but it is impossible in the cleaned skeleton 

 to judge of the amount of convexity that is possible. This seems 

 to be not so great in the case of the phocids, however. 



The transverse processes of the first 11 thoracics in Zalofhus consist 

 merely of blunt protuberances above the costal facets. They embody, 

 however, met-, an-, and possibly diapophyseal elements, which begin 

 to separate in the twelfth. Metapophyses are of increasing distinct- 

 ness in the thirteenth, fourteenth, and fifteenth, but anapophyses are 

 absent upon the fifteenth thoracic. There are certainly no diapo- 

 physes to the posterior thoracic vertebrae of Phoca and a small but 

 sharp anapophysis is present upon the fifteenth ; otherwise the lateral 

 details of the two genera are very similar. The first 10 ribs of Phoca 

 have both capitular and tubercular attachment to the vertebrae, and 

 of the remaining 5, capitular only. Because of the poorer definition 

 of the facets in the Zalophus skeleton (disarticulated) these details 

 could not be determined with certainty, but because of vertebral 

 similarity I judge that the costal conditions are the same. 



Lumbar vertebrae. — Unfortunately the Zalophus skeleton had but 

 four lumbar vertebrae and the one of Phoca hispida three. Thomp- 

 son (1909) ascribes to Barrett Hamilton a statement that "in most 

 seals the numbers of thoracic and lumbar vertebrae appear to be 

 usually 15 and 3, more rarely 14 and 6." The above figure 3 is 

 probably a misprint for 5, and with this exception I can find no 

 published statement of any otariid or phocid with less than five in 

 this series, while the number appears always to be six in Odohenus. 

 It therefore seems justifiable to assume that one lumbar vertebra 

 from the Zalophus and two from the Phoca have been lost and ta 

 compute the lumbar length on the basis of five vertebrae. In Zalophus 

 this computed item was 22 and in the Phoca 26 per cent of the body 

 length (46 in a cat) , and yet in proportion to general body mass the 

 lumbar series seems relatively the longer in Zalophus. On the whole 

 these vertebrae are of the same character as the more caudal of the 

 thoracic series save that there are no vestiges of anapophyses and the 

 inferior processes consist of broad parapophyses, relatively much 

 better developed in the phocid. In conformity with the previously 

 mentioned fact that in the latter the zygapophyses are much farther 



