332 AXXALS XEW YOBK ACADEMY OF SCIEXCES 



in primitive Ehipidistia but also in Dipt eras (Fig. 1. A ) a longitudinal 

 chain of small elements, bearing the supraorbital sensory line and extend- 

 ing back to the occiput. This series corresponds in position with the 

 postfrontals, intertemporals and suprasquamosals (pterotics or true supra- 

 temporals) of Stegocephali. 



The circumorbital series of Ehipidistia and primitive Dipnoi while 

 variable in number and form likewise correspond in position, and appear 

 to be collectively homologous with, the prefrontal. lacrA-mal. jugal. post- 

 orbital and postfrontal of Stegocephali. 



The spiracular cleft, which in Holoptycliius, Osteolepis microlepidotiLS 

 and Poh/pterus as well as in Dapedius is located behind and above the 

 orbit, appears to be lost in many Crossopterygii and in all Dipnoi, as 

 well as in the Stegocephali. 



The elements of the opercular series of Osteolepis and other Ehipidistia 

 appear to be only partly homologizable with those of Dipterus (Goodrich^ 

 1909, p. 2-iO) ; in the Stegocephali this series appears to have been lost 

 in part and in part preserved. In Osteolepis the preoperculum plus cheek 

 plate (X) was external to the quadrate : it was also beliind the postorbital 

 and jugal; and below the pterotic or true supratemporal : in all these 

 respects it corresponds perfectly with the lower temporal bone of Tetra- 

 poda. which is by many authors identified with the mammalian squa- 

 mosal. In Osteolepis and Diplopterus the preoperculum also bears a 

 sensor}' tract, as does also the squamosal of certain Stegocephali {e. g., 

 Ceraierpeton, Tutidanus, Moodie, 1908). In other Ehipidistia (e. g., 

 Tristicliopterus, Diplopterus, Megalichthys) the preoperculum has an 

 essentially identical position, but includes two elements, a posterior one 

 corresponding to the back part of the preoperculum of Polypterus and 

 to the whole of the preoperculum of Amia and an anterior one (X) 

 corresponding to the posterior row of postorbitals in jirimitive Holostei, 

 as suggested by Tracpiair. 



Below the preoperculum (squamosal) in Osteolepis microlepidotus, 

 Diploptenis, Megalichthijs. etc., are one or two elements {X'), together 

 corresponding in position to the interoperculum of higher fishes, which 

 have the topographic relations of the quadratojugal of Stegocephali and 

 like that element also bear a sensory tract. 



The retention and full development of the operculum, suboperculimi, 

 median paired and lateral gulars sharply separate the Eliipidistia from 

 the Stegocephali. The fate of these elements is discussed below (p. 337). 



The under side of the skull should, on general principles, ^deld impor- 

 tant evidence bearing on the problem of the relationships of the Tetra- 

 poda and the Ehipidistia. Unfortunately this resrion was not described 



