76 TRANSACTIONS OF THE AMERICAN PHILOSOPHICAL SOCIETY 



rising from its own center of ossification. There is reason to believe, however, that the 

 subdivision of the skull into separate bones has been conditioned chiefly by the necessities 

 of growth and nutrition and that originally the endocranium was a continuum and the 

 dermocranium consisted of a shell of ectosteal tissue, covering the chief functional regions 

 or organs. Even now after the separate bones have enjoyed many millions of years of 

 individuality, they are primarily regional subdivisions of functionally organic groups or 

 tracts as well as organs in themselves. 



In nearly all the hosts of typical fishes the syncranium is concerned with the pursuit 

 and capture of living prey, the exceptions being few and peculiar forms such as the parrot- 

 fishes and the like, which have given up this freely competitive roving life and become 

 highly specialized for living either on aquatic vegetation or on sessile animals. 



What then is the phylogenetic history of this peculiar and unique arrangement of 

 parts.'' There is strong evidence for inferring that even before the Ordovician period the 

 prevertebrates were bilaterally symmetrical, forwardly-moving animals in which the paired 

 olfactory, orbital and otic capsules were arranged in an invariable antero-posterior order 

 and in this fixed relation to the expanded oropharynx. The brain and cephalic sense 

 organs were protected by the connective tissue and tough skin of the head and roof of the 

 orobranchial chamber. The nervous system controlled the simple locomotor organs; these 

 consisted chiefly of bilateral rows of zig-zag muscle segments tapering to the rear and 

 causing undulations of the body as a whole. By means of its sense organs, nervous system 

 and locomotor apparatus the primitive chordate was able to move toward suitable sources 

 of energy, toward its mates and away from danger. The further inference is also highly 

 probable that long before the typical fish stage was reached, the vertebrates already sub- 

 sisted on living, moving prey, even if that prey was of quite small size. 



The ancient ostracoderms, or pre-fishes, are first known from a single plate found in 

 rocks of Middle Ordovician (Harding) age near Canyon City, Colorado. The class is 

 represented in the Upper Silurian of Spitzbergen, Oesel, Norway, Scotland and North 

 America by many genera belonging to several orders and to a considerable series of families. 

 In the succeeding Devonian period the ostracoderms gradually declined and for the most 

 part became extinct. The profound researches of Stensio (1927) and Kiser (1924) have left 

 no reasonable doubt however, that one or another of the ostracoderms gave rise to the 

 modern class of cyclostomes, including the lampreys and hags, thus confirming the earlier 

 views of Cope and others. 



The true or gnathostome fishes are not known until the Devonian period and even up 

 to the present time there are no known forms which definitely connect them with the 

 ostracoderms. Nevertheless there is a common fundamental plan possessed by cyclo- 

 stomes, ostracoderms and true fishes, especially in the arrangement of the parts of the 

 brain and spinal nerves and general anatomy, and as investigations multiply there is less 

 and less reason for doubting that the Agnatha (including ostracoderms and cyclostomes) 

 and the Gnathostomata (or true fishes and higher vertebrates) at least have a common 

 ancestral source, which would assuredly be nearer in most features to the agnathous than 

 to the gnathostome ground-plan. 



The earliest known gnathostomes, represented by the Palaeozoic sharks and ganoids 

 and their modern descendants, are already so far advanced in their adaptations for swift 

 swimming and predaceous habits that we necessarily seek for more primitive conditions in 

 earlier horizons. 



