100 TRANSACTIONS OF THE AMERICAN PHILOSOPHICAL SOCIETY 



In spite of its present specialization the lamprey skull retains the fundamental features 

 of the ostracoderm skull. Its otic capsule, for example, has but two semicircular canals, 

 like that of the cephalaspid ostracoderms and unlike those of all gnathostomes. 



The head of the Devonian antiarch Bothriolepis, which has been fully described by 

 Patten (1912) bears a rounded bony cephalic shield, which articulates with a bony thoracic 

 buckler. The shield and buckler are composed of numerous separate bones, many of which 

 appear to be homologous with those of the arthrodires (Hussakof, 1906, pp. 130-132). 

 According to Patten, small bony pieces around the mouth function as premaxillae, maxillae 

 and dentaries even though they may not be homologous with them. 



In the Devonian Macropetalichthys as described by Stensio (1925) we find a fish with a 

 shark-like brain and a well developed endo- and dermo-cranium. Some of the derm bones 

 are homologous with those of the arthrodires, others are not easy to identify. Olfactory, 

 optic and otic capsules are surrounded by a continuous matrix. Some of the dermal 

 exoskeletal plates bear channels for sensory canals. On the whole, Macropetalichthys 

 seems to be related to the elasmobranchs but to have advanced in the direction of the 

 ganoids without being ancestral to them. 



The cranial anatomy of the arthrodires has recently been very effectively dealt with by 

 Heintz (1931, '32), who has given an excellent restoration of the skull of Dinichthys. The 

 jaws of this fish, as described by Adams (1919) and by Heintz, were operated by a set of 

 muscles which apparently are wholly different from the jaw muscles of typical fish. The 

 old view that the arthrodires were related to the dipnoans now seems quite untenable. 

 They were rather a wholly extinct group, which nevertheless represented an abortive at- 

 tempt of a primitive gnathostome stock to rise from the elasmobranch to the teleostome 

 grade. While their olfactory, optic and otic capsules were doubtless homologous with 

 those of teleostomes, it seems probable that the individual bones of the dermocranium were 

 independently evolved in the two groups. The arthrodire jaws and jaw muscles also ap- 

 pear to have been independently evolved. 



Sharks, Rays, Chim^roids 



According to the evidence adduced by Stensio (1925, pp. 160-164; 187-189) it appears 

 that the cartilaginous condition of the skull in modern elasmobranchs is not improbably a 

 result of degeneration, as in the better known cases of the cartilaginous skulls of sturgeons, 

 spoonbills, Ceratodus, salmon, etc. Thus even the exoskeleton of modern sharks is retro- 

 gressive and now represented only by the skin and shagreen armor. In this connection it 

 must be admitted that in the oldest acanthodian sharks the exoskeleton was well developed 

 (Dean, 1907) and that, according to Reis (1896) and Jaekel (1927), the skulls of acanthodians 

 were much more primitive than those of modern sharks, showing subdivisions of the 

 palatoquadrate arch and of the Meckel's cartilage that correspond respectively with those 

 of the branchial arches. Dean however (1907) was skeptical as to the reality of these 

 subdivisions, while Goodrich (1909, p. 190) figures the palatoquadrate and Meckel's cartilage 

 of an acanthodian as severally undivided and essentially shark-like in form. 



By the time of the Upper Devonian cladodont sharks {Cladoselache) the exoskeleton 

 had already been weakened to a condition of delicate shagreen denticles and feeble teeth. 

 On the other hand, the Devonian genus Gemundina, recently redescribed by Broili (1930) 

 from excellent material, is a depressed form with wide pectorals, in which the exoskeleton 



