TERRESTRIAL VERTKRRATA. 251 



of the Biitrachia has been to the extinction of the centrum, the line of the Embolomeri 

 tends in the opposite direction, or towards the type of tlie Keptilia. Althou<;h the- 

 general characters of the skull of Cricotus are Bati-achiaii, the presence of a free pio- 

 atlas (Fig. o,n h) adds to the evidence of reptilian afHnity.* It is probable then that 

 we have in the' Embolomeri that order of Bati-achia from which tlie Kei)tilia were 

 derived, tluough intermediate forms not yet discovered.! It is also evident that the 

 Sphenosainid;\; cannot be referred to this order as T have proposed, but that they con- 

 stitute a family of lihachitomi. 



We have thus clearly shadowed forth in the Permian Vertebrata the ancestry of 

 the existing true Fishes,^ Batrachia, Keptilia, and Mammalia.§ 



Tir. — ALTERNATIVE EXPLANATIONS. 



I may here consider other possible ways of interpreting the homologies of the seg- 

 ments of the rhachitomous vertebral column, as follows: Let us first suppose with 

 Gaudry and Fritsch that the segments called intercentra in the figures and plates 

 accompanying this paper, are the true centra, and that the chevron bones are not 

 continna, but are originally separate, and have l)ecome coossified with the elements 

 with which we find them now contiinious. It will follow that the lai'ger bodies of the 

 vertebral column of Cricotus, Avhicli in Ihc dorsal region support the neural arch, are 

 intercentra, and that the centra are in llial genus in process of extinction. The same 

 will Ix' tiue of the so-called dorsal, and part or ail of the so-called caudal vertebra* of 

 Sphenodon. and of the Pelycosauriii. We thus have the Tleptilia in the |)Osition which 

 I have assigned to the Batrachia, the terms centrum and iutercentrum being mei'cly 

 reversed. As, however, these names were first a])j)lied to the Pelycosaurian reptiles 

 by me, and as it will recpiii-e less change of nomenclature to retain these names for the 

 bodies as they appear in Mammalia and Keptilia, it will be better to maintain the 

 proposition that the Batrachia have lost their centra, and retain only intercentra. 



Another alternative is to regard the hypocentrum ])leurale as the intercentruui w itli 

 the chevron luiuc ll will be remembered that the iiypoceuliiiiu pleurale is described b}' 

 Dr. Fritsch as lying below the i)leurocentra. This is of course on the anterior side of 

 the "hyi)ocentium.'' It looks reasonable to suppose that in the completion (»f (he body 

 of which the pleurocentrimi is a J)art, the hypocentinm pleurale should be included. 

 Su(h might be supposed to be the case in the cenfrum of Cricotus, and of llie Pel}'- 



* The iiniallanliil renlrilin is llinroiiglily ciiossific'il in all llio saliiiiiaiiiki" laiva- aliovc rcferiiil lo. 

 t I have suggested this view in ilie American Naturalist, \HSi, p. ;i7 ; \>i8't, p. 77. • 



X .\<l(litioDa1 specimens of Didjmodus display very distiiiclly the futures I liave described, wliidi >Ir. Onrmnn 

 tldnks to be nccidenlal. 



S A second posterior fool of C'lcpsydrops displays perfectly the iii:iminali;in characters I lijivc ascribed to It. 



A. V. S. — VOL. \VI. 2f. 



