• 1It has been assumed that, in accordance with experimental evidence, the initial patterns of dermal bones result from the relative positions and manner of growth of their rudiments.
  • 2The sutural patterns resulting from freely growing bones have been examined for different initial arrangements, even or elliptical bone growth, and uniform or uneven growth of the embedding dermis.
  • 3It has been shown that the bones of the “parietal shield” of osteolepid fishes form a pattern consistent with elliptical bone growth according to the known centres of ossification, and that when additional interparietal bones are present the suture pattern conforms to the same assumption.
  • 4It has been shown that the bones of the “parietal shield” of holoptychiid fishes also conform to the assumption of elliptical bone growth, and differ from those of the osteolepids only in the more anterior position of the centres of ossification of the interparietals.
  • 5The sutural. pattern of the temporal region of normal labyrinthodonts has been shown to conform to the assumption of elliptical bone growth, and the absence of an intertemporal bone in later forms is consistent with its loss and the invasion of its territory by the prefrontal and supratemporal bones.
  • 6The occupation of the region of the intertemporal bone by a lappet formed from the parietal is shown to be consistent with the loss of the bone in the presence of evenly growing neighbouring bones. It is pointed out that this occurred apparently in holoptychiid fishes, Elpistostege, the Ichthyostegalia, and in reptiles.
  • 7It is shown that the reduction of the supratemporal bone to a splinter is consistent with the assumption of its appearance only at a late stage of growth of the parietal.
  • 8It is claimed that since the outstanding variations in the dermal bone patterns of early vertebrates can be accounted for quite closely by simple assumptions concerning the manner of growth and relative positions of their rudiments, assumptions of bone fusions without direct evidence are unjustified.
  • 9It is pointed out that the analyses of the “parietal shields” of osteolepids and holoptychiids by the methods adopted in this account imply that lateral line canals are not constant in their relations to dermal bones.
  • 10It is suggested that the examination of dermal bone patterns with their probable manner of development in mind emphasizes similarities and differences which are likely to be of taxonomic value.