AQUATIC MAMMALS 



that it will henceforth play a role of increasing unimportance. Inci- 

 dentally any function of steering which it may now have would be insuf- 

 ficent to save it from atrophy, in the latter course of events, for it is situ- 

 ated too close to the primary propulsive organ to act efficiently as an 

 equilibrator. 



The tail in the Sirenia varies. In Haltcore and Hydrodamatis it is 

 quite whale-like, there being two pointed flukes with a partly denned 

 medial notch between, and a constricted peduncle. Little more than 

 this can be said, for the muscular anatomy of neither has been inves- 

 tigated; but in main features, consisting chiefly of simplification of the 

 apaxial, and fusion and hypertrophy of hypaxial elements, it undoubt- 

 edly resembles the Cetacea. In Trichechus the tail is usually stated as 

 shovel-shaped, but in its exact conformation there is probably specific 

 variation. Thus Murie (1872) figured a specimen (figure 6) of what 

 he calls Manatus americanus ( = T. latirostris) from the West Indies 

 in which there is a medial indentation of the posterior tail, while else- 

 where (1880) he showed an individual from British Guiana with tail 

 tip somewhat pointed (figure 6) . But even yet it is not known whether 

 the animals from these two regions really represent two species. The 

 illustrations will give a better idea of the sirenian tail than can a de- 

 scription. It will be noted that in this respect the manati is much less 

 specialized than the dugong, but that the former represents a stage 

 through which it is not improbable that the latter at one time passed. 

 In Trichechus there is no well defined peduncle, but merely a slight 

 taper of the posterior part before the lateral expansions of the tail be- 

 gin. The latter is readily seen to be fairly intermediate between a stage 

 on the one hand wherein there was either no lateral broadening of the 

 tail, or one comparable in degree to that existing in Enhydra, and on the 

 other, conditions as now to be found in Haltcore. In a mounted skele- 

 ton of Trichechus in the National Museum there are 22 lumbo-caudal 

 vertebrae, while in a specimen of Haltcore this series numbers 26. In 

 the latter the transverse processes gradually diminish in width from the 

 thorax to the region of the peduncle, thence widening once more until 

 near the caudal tip. This is an interesting occurrence for the reason that 

 it is a character which the Cetacea do not show to the slightest degree. 

 It seems to be an additional instance of the fact that two animals fre- 

 quently respond differently to the same stimulus. 



The caudal conditions in the Sirenia are not readily analyzed. It is 

 currently believed that the sirenian ancestor was of proboscidean stock 

 and we would therefore picture the caudal stimuli to have been some- 



[196] 



