GREGORY, PRESENT STATUS OF ORIGIN OF TETRAPODA 



323 



the mouth and on the inner side of the mandible and the loss of marginal 

 teeth on the premaxill^e, maxillae and dentaries. 



The earliest Dipnoi have "paralleled" the Tetrapoda and other pro- 

 gressive types in that the preorbital rostrum is expanded and the orbits 

 are relatively far posterior though still of small size ; the pattern of the 

 skull-roof, with its paired "frontals," "parietals" and other elements, is 

 also probably analogous, rather than homologous, with that of Tetrapoda. 



Our conception of the relationship of the Dipnoi to the Tetrapoda is 



Fig. 1. — Pattern of sJcull-top of Devonian dipnoans 



A, Dipterus, after Goodrich, slightly modified; B, Scaumenacia curta, after Hussakof. 



In specimens of Dipterus the numerous sensorj'^ pits are scattered over broad tracts, 

 the general directions of which are indicated hy the dotted lines, except in the occipital 

 region where the dotted lines represent shallow grooves. The principal sensory tracts 

 are in general similar to those of Stegocephali. 



The "parietals," "frontals," etc., are probably analogous rather than homogenous with 

 those of Tetrapoda. 



Dso, dermosupraoccipital ; S.t, supratemporal (pterotic) ; Th^ tabulare (epiotic) ; F.p, 

 preparietal ; Fr, frontals ; Na, naso-ethmoid region. 



to some extent dependent upon the validity of Dollo's view (1895) that 

 Dipterus is the most primitive known Dipnoan, structurally ancestral to 

 all the later types. The skull top of the modern Ceratodus is so widely 

 different from that of Dipterus that Dr. Eastman (1907, p. 95) lias con- 

 cluded that the two forms have no near relationships with each other and 

 that the living genus represents some other and independent line that 

 has come doA\Ti from a pre-dipterine stock. But after comparing the 

 skull patterns of Ceratodus, Ctenodus, Phaneropleuron and Scaumenacia 



