420 Br. Si, Woodward — On the Fossil 8irenia. 



The atlas- and axis-vertebice in Uliytina are fairly robust, and the 

 atlas is as broad as the second dorsal vertebra ; but the five re- 

 maining cervical vertebra, although quite free, are thin and plate- 

 like, as in the Cetacea proper. But the Sirenia are distinguished 

 from the true Whales by their capability of moving the head from 

 side to side, and up and down, by means of the " odontoid process " 

 of the axis vertebra on which the head rotates. In the Cetacea, in 

 which the cervical vertebrse are anchylosed together to a greater or 

 less extent, and the neck is consequently immovable, the odontoid 

 jDrocess is also wanting. 



As the Sirenia spend their whole lives browsing upon the Lami- 

 narice and other Alg^ and aquatic plants, this power to move the 

 short neck pretty freely must be essential to them both in feeding 

 and also in putting up their heads to breathe. 



The number of vertebras attributed to the Sirenia, both of living 

 and extinct genera, is very variable according to different authors. 

 Prof. Brandt attributed to BJiytina 7 cervicals, 19 thoracic or dorsal 

 vertebrge, and from 34 to 37 lumbar, sacral, and caudal. The 

 cervicals and dorsals are readily determined ; but, as none of the 

 vertebree are anchylosed together to form a sacrum, it is a matter of 

 some difficulty to decide which are lumbar and which are sacral 

 vertebra. 



Not only does anchylosis never occur in the vertebrae of the Sirenia, 

 but the flat ends of the centra of the vertebrge do not ossify separately 

 so as to form dish-like epiphyses in the young state, as is commonly 

 the case in all the other Mammalia. 



Brandt indicates the 7th vertebra beyond the last of the dorsal 

 or thoracic series as bearing the rudimentary pelvis ; but as the 

 vertebree are never anchylosed to form a sacrum, we can onl}' con- 

 jecture (by noticing a slight prominence upon the posterior border 

 of the extremely wide transverse processes) which of these lumbar- 

 sacral vertebrge seem marked as sacral, probably about the 5th, 6th, 

 and 7th. The 13 vertebree next behind the dorsal series may, from 

 their size and their wider and longer transverse processes, be con- 

 sidered lumbar and sacral, and the 21 following vertebrae as caudal ; 

 about 6 or 8 of the most anterior of the latter had small yshaped 

 chevron bones or htemal arches attached to them in the St. Petersburg 

 specimen. 



The transverse processes in the caudal series are much smaller, 

 thicker, and shorter, and are directed obliquely backwards. 



There is a marked variation in the form and size of the neural 

 arch and the centrum of the several vertebrae in the spinal column, 

 from before backwards. The anterior dorsal vertel)r8e have each 

 a small compressed centrum, much broader than deep; the neural 

 arch is triangular, the neural spine erect. 



From the 5th to the 8th dorsal the centra are longer and cordiform, 

 and the neural canal is smaller and more rounded : the neural spine 

 bends backwards, and the zygapophyses are more prominent. 



The lumbar vertebrge are much dilated laterally, the centra being 

 nearly three times as broad as deep. The neural canal is reduced 



