SKELETON OF LEPIUOSTEUS. 455 



caudal vertebi'se each centrum is divided into a prsecentrum 

 (without arches) and a postcentrum. 'i'he terminal vertebras, 

 hypm-als, etc., are arranged much as in Lepidosteus. The 

 skeleton of the fins is as in Lepidosteus, except that the pectoral 

 radials are more numerous and more than one articulates 

 with the metapterygium ; the pectoral arch differs from that of 

 Lepidosteus in that the coracoid cartilage is unossified and the 

 post-temporal is large and loosely attached, with a lower foi-k 

 running to the opisthotic. 



Systematic Position of Lepidostexis and Amia. 



It is probable that most of the characters in which Lepidosteus 

 s^ndi Amia agree are common to all the Holostean Ganoids (Semio- 

 notidse, Pycnodontidse *, Macrosemiidae, Eugnathidae, Amiidfe, 

 Pachycormida? f, Aspidorhynchidps). Further, it is evident that 

 the essential features which distinguish Lepidosteus from Amia, 

 i. e. the characters of the jaws, suspensorium, and opercles, 

 distinguish it also from all the others. Lepidotus is extremely 

 like Amia in its head skeleton, except that the skull is more 

 compressed, with the epiotics meeting above the exoccipitals and 

 the orbitosphenoids imited {cf. Smith Woodward, P. Z. S. 1893, 

 and Mon. Palseontogr. Soc. 1916-1919). I am fairly certain 

 that I can make out in this genus the symplectic articulation 

 with the lower jaw. 



The Semionotidse, to which family Lepidotus belongs, first 

 made their appearance in the Upper Permian, and are very 

 distinct in structure from the Palseoniscidse, the only fishes 

 known which can be regarded as their ancestors. The typical 

 PalsBoniscids appear to have been swift-swimming predacious 

 fishes, with a lai^ge mouth and sharp teeth. If we regard them 

 as giving rise to the Semionotidse, probably slow-swimming 

 bottom-feeding fishes, with a small mouth and styliform or 

 tritoral teeth, we can interpret many of the differences as related 

 to a change of habits. 



The change from dorsal and anal fins with numerous rays 

 forming a close-set series to fins with the rays relatively few and 

 spaced, each ray with its own pterygiophore, and the reduction 

 of the muscular lobe at the base of each fin and of the radial 

 segments of the pterygiophores, are readily understandable if 

 the use of these fins changed from cleaving the water and with- 

 standing strains to performing the delicate movements of a fish at 

 rest or swimming slowly. Correlated with this modification of the 

 dorsal and anal fins is that of the caudal, the upturned end of the 



* These are aberrant in the reduction of the opercular apparatus and the arrange- 

 ment of the bones of the cranial roof. 



t The praBmaxillaries are said to be loose and separated by an ethmoidal rostrum, 

 but 1 am not satisfied that this interpretation is correct. I think that the so-called 

 praemasillaries may be the fractured anterior ends of the maxillaries. 



30* 



