GREGORY. PRESEXT STATUS OF ORIGIN OF TETRAPODA 327 



rangements of the sutures which produce the distinctive skull patterns 

 of these early types. It was recognized by Kyder (1892) that the shape 

 and arrangement of the scales of fishes was originally a result of the move- 

 ments of the myomeres^ but so far as I know it has not hitherto been 

 taught that the shape and arrangements of the dermocranial elements^ 

 which in primitive fishes are histologically homologous with scales, were 

 likewise the result in part of muscular strains and stresses. It may be 

 stated as a general hypothesis that in the dermocranium of primitive 

 fishes the position and arrangement of the sutures and the consequent 

 pattern of the osseous ''elements" are the evolutionary resultants of the 

 various symmetrically balanced stresses induced by the action of the 

 underlying muscles of the eyes, jaws, hranchial arches and pectoral limhs. 

 in composition luith the position and size of the olfactory, optic and 

 auditory capsules. It is at least a fact that sutures and other articula- 

 tions define loci of relative mobility, centers of ossification define loci of 

 relative stability. Differential growth of one region of the skull, as in the 

 rapid elongation of the snout, also results in more or less rearrangement 

 of the sutures and osseous elements. 



I would also advance the hypothesis that the whole dermoskeleton of 

 the head in the unknown pre-Devonian forerunners of the Dipnoi. Cros- 

 sopterygii and Actinopterygii was a continuous membranous covering, 

 consisting (see Goodrich, 1909, pp. 215-217) of minute cosmine tuber- 

 cles underlain by imperfectly differentiated vasodentine and isopedine 

 and entirely homologous with the membranous covering of the trunk and 

 fin-folds. Wherever movements of this membranous covering were pro- 

 nounced, as in the opercular region, around the orbits and near the spi- 

 racular clefts, there Avere deep creases, but where movements were less 

 pronounced the creases were proportionally shallow. When the critical 

 stage of ossification began, in the forerunners of the Dipnoi, Crossop- 

 terygii, Actinopterygii, the roof of the dermocranium commenced inde- 

 pendently in the several lines a process of fragmentation; so that the 

 rostrofrontal segment was split up into premaxillae, nasals, ethmoids, 

 frontals ; and the orbito-parietal segment was divided into its component 

 "elements.'' 



The excessive subdivision of the maxillary and preoperculum of Lepi- 

 dosteus, of the parafrontals of coelacanths, of the spiracular plates of 

 Polypterus furnish extreme examples of fragmentation. In the skull- 

 roof of Osteolepis, on the other hand, we may have an example of a form 

 in which ossification of the dermocranium had already taken place, ])iit 

 fragmentation of the "cranial buckler" was still incomplete. In some 

 specimens of Osteolepis the median suture between the frontals extends 



