PHYLOGENY. 91 



The Batrachia may then have originated from a hyo- 

 stylic teleostomous (i. e., one with complete maxillary 

 arch) fish. Among Teleostomata we naturally look for 

 forms with limbs which approach nearest the batrachian 

 type, and in which median fins are feeble or wanting. 

 Such are the Rhipidopteryg:ia (formerly included in 

 the Crossopterygia), which include the families of 

 Holoptychiidae, Tristichopteridae, Osteolepididae, Coe- 

 lacanthidae, and perhaps some others. These fam- 

 ilies, except the last, abounded in. the waters of the 

 Devonian period, at the time when the ancestors of 

 the Batrachia also existed. All of them agree in pos- 

 sessing the median fins of greatly reduced proportions, 

 and the mesodermal or internal elements of the paired 

 fins more like the limbs of the Batrachia than are those 

 of any known fishes. The constitution of the superior 

 cranial wall is a good deal like that of the stegocepha- 

 lous Batrachia. The characters of the fins can be 

 learned from the accompanying figure of the Eusthe- 

 nopterumfoordii'Whitea.-ves, one of the Tristichopteridae. 

 The pectoral fin well-nigh realizes Gegenbaur's theory 

 of the derivation of the Chiropterygium from the Ar- 

 chipterygium. 



The ancestral type of fishes is probably the acan- 

 thodean order of the subclass of sharks (Elasmo- 

 branchii).! Like other sharks, they are hyostylic and 

 have no maxillary arch or cranial bones. They have 

 the ptychopterygium, which is the primitive type of 

 fin. In this fin the osseous elements which support 

 the fin-rays are enclosed within the body-wall, the rays 

 only being free. Such a fin sustains the hypothesis 

 that the paired fins are parts of primitively continuous 



1 As represented by the Cladodontidae ; see Dean, Trans. N. Y. Acad. 

 Sci., 1893. p. 124, and Cope, Afnerican Naturalist , 1893, p. 999. 



