Geological Society of London. 423 



are described in detail, and the characters of the genus are compared 

 with those presented by other genera of Sirenians, both living and 

 fossil, especially Manatas and Felsinotherium. The dental formula 

 of Prorastomus is given as : — 



. 3—3 , 1—1 5-5 3—3 . 



i. 3 - 3? -, d. or c. j-j, p. 5= 5 , m. 3 _ 3 = 48 ; 



thus, as in Manatus, showing an excess in the molar series over the 

 type of the terrestrial herbivorous mammalia, whilst the incisors and 

 canines retain the common type as to number and kind, and have 

 not been subjected to so great a degree of suppression or of indivi- 

 dual excess of development as in existing Sirenians. The presence 

 of these small subequal incisors in both jaws of Prorastomus is the 

 most marked feature in which Prorastomus adheres to the normal 

 mammalian type, while showing the essential characters of the 

 marine Herbivores ; but a similar tendency is shown in other parts 

 of the skull. 



The author regards the Sirenia as essentially monophyodont. 

 Halicore and Felsinotlierium depart further from the type than 

 Halitherium and Manatus, and these than Prorastomus. Bhytina, 

 with a better developed brain, and with the jaws edentulous when 

 adult, is an extreme modification of the Sirenian type. The rudi- 

 mentary femur in Halitherium is to be regarded as the result of 

 degeneration through lack of use, from better-limbed prototypal 

 mammals. 



With respect to the genealogy of the Sirenia, the author remarks 

 that Hackel derives the Sirenia, Zeuglodontes, and Cetacea, toge- 

 ther with the Artiodactyla, from the branch Ungulata, and the 

 Perissodactyla from the branch Pycnoderma of the Mammalian 

 trunk ; but that while Halitherium and Felsinotherium show the 

 molar pattern of Hippopotamus, Prorastomus exhibits that of 

 LopJiiodon and Tapirus, to which Manatus also adheres, rather than 

 to any Artiodactyle type. The author suggests that both Ungu- 

 lates and Sirenians diverged at some remote period from a more 

 generalized (Cretaceous ?) mammalian gyrencephalous type ; and 

 that the marine Herbivora in the course of -long Eocene and 

 Miocene eons were subjected to conditions producing modifica- 

 tions of their molars, leading on one side to an Artiodactyle and 

 on the other to a Perissodactyle character. As Prorastomus by its 

 more generalized dentition and shape of brain represents a step 

 nearer the speculative starting-point than any other Sirenian, it 

 acquires a great interest, and the determination of the precise age 

 of the (supposed Eocene) bed from which its remains were derived 

 is very much to be desired. 



2. " On the Structure of the Skull of Bhizodus:' By L. C. 

 Miall, Esq., F.G.S. 



In this paper the author described a large skull of Bhizodus from 

 the coal-shale of Gilmerton, near Edinburgh. The characters de- 

 scribed show that Bhizodus is a Ganoid fish, and that its position in 

 the order is not far from Holoptychius and Megalichthys. The author 

 referred it to the cycloidal division of the family Glyptodipterini. 



