336 



ANNALS NEW YORK AGADE2IY OF SCIENCES 



The supposed correspondence in the elements of the lower jaw between 

 the Osteolepidffi and Khizodontidag on the one hand and the most primi- 

 tive Stegocephali on the other have been noted above (p. 334). 



According to Watson, the resemblances in the under side of the skull 

 between Megalichthys and the primitive Stegocephali {Loxomma, etc.), 

 which have a single occipital condyle, is remarkably close, and apparently 

 there is little doubt as to the homology of the following elements in the 

 two groups: premaxillae, maxillae, prevomers, palatopterygoids, para- 

 sphenoid, quadrates, hyomandibular (stapes), basisphenoid, basioccipital 



Fig. 3. — Skull patterns of Trimerorhachis and Diplopterus 



A, Permo-Carboniferous stegocephalian, Trimerorhachis mediits (order Temnospondyli), 

 after Broom. B, Devonian osteolepid rhipidistian Diplopterus, after Jaekel (lettering 

 somewhat altered). 



Abbreviations as in Fig. 2, p. 330 ; also : It, intertemporal ; H.m, hyomandibular (colu- 

 mella auris) ; ? sy, ? symplectic (? extra columella). 



In B the interfrontal foramen is identified by authors as the pineal opening. The 

 brain M-as probably located far forward. 



(a modified centrum). Again the elements of the chondro cranium, so 

 far as they are known in Stegocephali (Trimerorhachis, Eryops), seem 

 to correspond in general plan with the conditions in the existing 

 Polypterus. 



It is conceivable that the Stegocephali may have branched off from 

 the stem of the Ehipidistia before the skull-roof became fully ossified, 

 but the existence of such an extensive series of correspondences offers 

 strong evidence of community of origin. It seems reasonable, therefore, 

 to conclude that the primitive stegocephalian skull has been derived from 

 the rhipidistian skull through the following advances : 



