14 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 70 



what caudad. Due apparently to the incomplete fusion of the two 

 halves of the vertebral arch as these develop from their respective 

 centers of ossification, some of the thoracic spines (see table, p. 11) 

 in all of the skeletons save 49544 are partially bifid. 



Lurribar vertebrae. — The lumbar vertebrae are usually of such 

 number as to compose with the thoracic series a complement of 25. 

 Allen lists only two individuals with 25, one with 27, and the remain- 

 der with 26, while but one of my specimens is in the first-mentioned 

 class. I imagine, however, that the vertebra which Allen considers 

 to be the last lumbar I have placed in the caudal series. The length 

 of the lumbar vertebrae varies from 25 to 29 per cent of the total 

 vertebral length. The diapophyses of the first lumbar are abruptly 

 of considerably greater length than those of the last thoracic. The 

 third, fourth, or fifth of this series is the broadest (the fourth varies 

 from 102 to 158 mm.), and thence caudad to the end of the tail there 

 is a gradual decrease in the length of the diapophyses. The meta- 

 pophyses are as described under the thoracic vertebrae, and there 

 are no other processes except the neural spines. The more cranial of 

 these are sometimes bifid, but not to the extent found in the thoracic 

 series. In 239990 and 49544 the lumbar spines increase in height 

 caudad, but in the others they are all approximately of the same 

 size and but a trifle, if any, highier than in the thoracic region. 



Caudal vertebrae. — There are 32 caudals in one, 31 in three, and 

 29 in one of the specimens before me. About a dozen of the ter- 

 minal vertebrae are missing from the tail of 240001, and by com- 

 parison it is clear that there are three absent from that of 240003. 

 Allen states that there are but 26 of this series in 49544, but he over- 

 looked, or else did not have, four (placing the fifth with the lumbar 

 series), and there are really 31, all unquestionably belonging to this 

 one specimen, for its bones are darker and greasier than any of the 

 others. The writer herein follows what seems to be the usual pro- 

 cedure in studies of cetacean osteology in considering that the first 

 vertebra with facets for the attachment of a chevron constitutes 

 the first of the caudal series. This, of course, is arbitrary, for there 

 is no way of knowing where the cetacean tail actually starts, as in 

 other mammals chevrons do not always begin at the same point in 

 the series. Allen evidently considers that the first vertebra showing 

 chevron facets is the last lumbar, although he does not actually say 

 so. In the only three instances where this could be measured with 

 accuracy the caudal series constitutes from 42 to 46 per cent of the 

 vertebral length. 



The more anterior of the caudal spines are the highest of the whole 

 vertebral column, but thence caudad all processes decrease in size in 

 caudal sequence until the terminal dozen are nothing more than bony 



