566 PROF. OWEN ON PRORASTOMUS SIRENOIDES. 



In the number of teeth developed in the molar series (PL XXIX. 

 fig. 3), Prorastomus departs, as already observed, from the diphy- 

 odont or higher mammalian type, yet only by a single additional 

 tooth, so far as can be made ont from the present fossil. In this, how- 

 ever, there are, in advance of the last three grinders, which from their 

 size and structure I have symbolized as true molars, five smaller 

 and simpler teeth, which I have described under the name of pre- 

 molars. "Whether any of these have displaced, vertically, deciduous 

 predecessors cannot be determined in the present unique evidence of 

 the genus*. I incline to think not, and to regard them as answering 

 to the series of teeth which are displaced by permanent premolars in 

 the diphyodont group. In the additional molar tooth of Prorastomus 

 we have the last show of " irrelative repetition" which marks the 

 dental series in the DerphinidaB, Zeuglodontidse, loricate JBruta, mar- 

 supial AmpMtheria, Myrmecobius, &c. 



The Sirenia are essentially monophyodont. The seeming excep- 

 tion in the upper incisors of Halicore, where a small functionless 

 tooth precedes the incisor tuskf, and is shed, leaving for some 

 time traces of its small socket anterior to that of the large tusk, 

 may support the interpretation that such tusk is a functionally deve- 

 loped second incisor, the homologue of d i 2 in Manatus, and not of 

 i 1 in Diphyodonts, preceded by a d i 1. 



Halicore and Felsinotherium depart further from type than do 

 Halitherium and Manatus, but these still further than did Prora- 

 stomus. Phytina, by its better-developed brain than in Prorastomus 

 and Eotherium, and by its adult edentulous condition of jaws, ex- 

 hibits a still more extreme modification of the Sirenian type. 



In Halitherium, besides the pelvic ossicles, a rudimentary femur was 

 articulated with an acetabulum, but remained, as in Baicena, hidden 

 under the skin. 



I infer, as in the case of the similarly concealed metapodials II. 

 and IV. in pliocene and existing Equidse, that such rudiments of 

 the hind limbs are the result of degeneration, through lack of use 

 or need, from better-limbed prototypal mammals. Hyracotherium, 

 PliolopJius, MiolojoTius, Coryphodon, from the older Eocene deposits, 

 indicate or exemplify a more generalized ungulate type than the 

 hoofed fossils from subsequent Tertiaries, and oppose the idea that 

 the later, more modified, better-defined Artio- and Perissodactyles 

 were derived from distinct primary branches of the mammalian stem. 



Prof. Hackel % derives Sirenia as follows in his ideal "genealogical 

 tree " of Mammals : — 

 " $TKE$iK=Pliycoceta 



- from the branch ' Ungulata ' 



Zeuglodontes = Zeugloceta 

 Q^TA.CEK=Autoceta .... 



Artiodactyla 



Perissodactyla from the branch ' Pycnoderma ' of 



the Mammalian Trunk." 



* They are so symbolized in PI. XVIII. fig. 1. 



t Proceedings of the Zoological Society, 1 838, p. 40 ; ' Odontography,' 4to, 

 1540, p. 384, pis. 92-95; Cyclopedia of Anatomy, art. "Teeth," vol. iv. (1852), 

 p. 902, fig. 575. 



| ' Generelle Morphologie der Organismen,' Bd. ii. tab. yiii. 



