AQUATIC MAMMALS 



three costal spaces. Hence, while the apaxial muscles operate through- 

 out the entire length of the vertebral column, the hypaxials can bend 

 only the lumbo-caudal series. Anteriorly what downward bending of the 

 posterior thorax and of the head is necessary in swimming must be taken 

 over chiefly by the powerful rectus abdominis on the one hand, and the 

 ventral neck muscles on the other. In this connection a point should be 

 mentioned that may already have occurred to the reader. In theory the 

 swimming motion of whales has been discussed as though the curve as- 

 sumed by the body were a perfect arc. In practice, however, this may not 

 be so. Not only may the costal equipment prevent as much possible cur- 

 vature of the thoracic vertebrae as of the lumbo-caudal series, but this 

 may further be reduced by the greater mass in the former region. It 

 might therefore be more correct to consider that the neck and tail each 

 bend from the thorax to a considerable degree, while the curve of the 

 thoracic vertebrae is more moderate. This, at the present time, is almost 

 impossible of determination. 



Theoretically lateral motions by the tail are unimportant to the Cetacea, 

 ■ but actually there must be considerable strength in this plane for thrash- 

 ing about when the need arises. Such movement may be accomplished 

 chiefly by the flexion of the apaxial and hypaxial muscles of a single side, 

 and also by the intertransversarii, which occupy the space between the 

 transverse processes of the lumbo-caudal vertebrae. In cetaceans these 

 are unusually modified and upon the thorax, of at least most odontocetes, 

 spread out into a thin sheet covering a large part of the ribs deep to the 

 latissimus dorsi layer. In Monodon there is a convergence of the fibers 

 from each direction to (about) the sixteenth lumbo-caudal vertebra, in- 

 dicating a center in this region for lateral movement. It is likely that the 

 intertransversarii act fully as much in preventing too much curvature as in 

 instigating it. In other words the function is probably as much static as 

 active. 



[182 1 



