318 ANNALS NEW YORK ACADEMY OF SCIENCES 



paired limbs are still ineontestibly ichthyoptervgia : and there are almost 

 equally great differences in the skull. 



Baur's view that the Amphibia and higher classes liave been derived 

 from Palaeozoic Crossopterygii has of late years been supported especially 

 by Watson (1912) and Broom (1913). Jaekel (1896, 1909), who de- 

 lights in such phylogenetic paradoxes, holds that the earliest Tetrapoda, 

 instead of havino' been derived from the fishes, have aiven rise to them 

 through progressive secondary adaptation to aquatic life. 



Although the true linking forms between fishes and amphibia still 

 remain unknown and must be sought in formations of Lower Devonian 

 or even Upper Silurian age (Moodie), comparative anatomy alone ofiiers 

 sufficient evidence to prove that the recent fishes and Tetrapoda, in spite 

 of their admittedly wide differences and remote separation, yet trace their 

 origin to a common source. One need cite only the fundamental unity 

 of plan that runs throughout the vertebrate series and the weighty char- 

 acters that unite all the Craniata into a natural group. One recalls, for 

 example, that creatures so diverse in form and mode of life as fishes, 

 amphibians, reptiles, birds and mammals all have an identical arrange- 

 ment of the six eye-muscles in relation to the eyes and to three of the 

 cranial nerves ; that the complex head is a synthesis of sensory capsules, 

 brain-trough, occipito-vertebral segments, primary jaws, branchial arches 

 and dermal elements ; that the whole locomotive and muscular apparatus 

 of trunk, limbs and head is evolved out of metamerically arranged 

 myomeres of identical embryonic history. Thus, although the differences 

 between known Amphibia and known fishes are very marked, they are 

 morphologically of far less weight than the resemblances. 



The main conclusion of the present paper is that among the various 

 groups of fishes which are known from the Devonian and Upper Silurian, 

 onl}^ one, the rhipidistian Crossopterygii, belongs near the ancestral line 

 or lines of the Tetrapoda. 



OSTRACODERMI^ AXTIARCHI AXD ARTHRODTRA 



As the Silurian and Devonian Ostracodermi include the oldest known 

 types of chordate animals it is pertinent to inquire what their relations 

 may be to the remote ancestors of the Tetrapoda. All the laiown ostraco- 

 derms appear to be aberrantly specialized in certain directions, but long 

 consideration of their many peculiar characters has convinced me that 

 they stand far below the true fishes and that the group as a whole may 

 represent an important stage in the genealog}' of the vertebrates. 



Of the many beautifully preserved specimens of Birhenia. Lanarhia, 

 Drepanaspis, Pteraspis, Tremataspis, Cephalaspis and allied genera, none 



