"1. Homologies of parts are best determinable, _ceteris paribus_, in the most nearly related forms. "2. Identification should proceed from a central or determinate point outwards. "The applications of these principles are embodied in the following conclusions: "1. The forms that are best comparable and that are most nearly related to each other are the Dipnoi, an order of fishes at present represented by _Lepidosiren_, _Protopterus_, and _Ceratodus_, and the Batrachians as represented by the _Ganocephala_, Salamanders, and Salamander-like animals. "2. The articulation of the anterior member with the shoulder-girdle forms the most obvious and determinable point for comparison in the representatives of the respective classes. [Illustration: FIG. 71.--Shoulder-girdle of _Amia calva_ (Linnæus).] [Illustration: FIG. 72.--Shoulder-girdle of a Sea Catfish, _Selenaspis dowi_.] =The Girdle in Dipnoans.=--"The proximal element of the anterior limb in the Dipnoi has almost by common consent been regarded as homologous with the _humerus_ of the higher vertebrates. "The humerus of Urodele Batrachians, as well as the extinct Ganocephala and Labyrinthodontia, is articulated chiefly with the coracoid. Therefore the element of the shoulder-girdle with which the humerus of the Dipnoi is articulated must also be regarded as the _coracoid_ (subject to the proviso hereinafter stated), unless some specific evidence can be shown to the contrary. No such evidence has been produced. "The scapula in the Urodele and other Batrachians is entirely or almost wholly excluded from the glenoid foramen, and above the coracoid. Therefore the corresponding element in Dipnoi must be the _scapula_. "The other elements must be determined by their relation to the preceding, or to those parts from or in connection with which they originate. All those elements in _immediate_ connection with the pectoral fin and the scapula must be homologous as a whole with the coraco-scapular plate of the Batrachians; that is, it is infinitely more probable that they represent, as a whole or as dismemberments therefrom, the coraco-scapular element than that they independently originated. But the homogeneity of that coraco-scapular element forbids the identification of the several elements of the fish's shoulder-girdle with regions of the Batrachian's coraco-scapular plate. [Illustration: FIG. 73.--Clavicles of a Sea Catfish, _Selenaspis dowi_ (Gill).] "And it is equally impossible to identify the fish's elements with those of the higher reptiles or other vertebrates which have developed from the Batrachians. The elements in the shoulder-girdles of the distantly separated classes _may_ be (to use the terms introduced by Dr. Lankester) homoplastic, but they _are not_ homogenetic. Therefore they must be named accordingly. The element of the Dipnoan's shoulder-girdle, continuous downward from the scapula, and to which the coracoid is closely applied, may be named _ectocoracoid_.