"Neither the scapula in Batrachians nor the cartilaginous extension thereof, designated suprascapula, is dissevered from the coracoid. Therefore there is an _a priori_ improbability against the homology with the scapula of any part having a distant and merely ligamentous connection with the humerus-bearing element. Consequently, as an element better representing the scapula exists, the element named scapula (by Owen, Günther, etc.) cannot be the homologue of the scapula of Batrachians. On the other hand, its more intimate relations with the skull and the mode of development indicate that it is rather an element originating and developed in more intimate connection with the skull. It may therefore be considered, with Parker, as a _post-temporal_. "The shoulder-girdle in the Dipnoi is connected by an azygous differentiated cartilage, swollen backwards. It is more probable that this is the homologue of the _sternum_ of Batrachians, and that in the latter that element has been still more differentiated and specialized than that it should have originated _de novo_ from an independently developed nucleus." [Illustration: FIG. 74.--Shoulder-girdle of a Batfish, _Ogcocephalus radiatus_ (Mitchill).] =The Girdle in Fishes Other than Dipnoans.=--"Proceeding from the basis now obtained, a comparative examination of other types of fishes successively removed by their affinities from the Lepidosirenids may be instituted. "With the humerus of the Dipnoans, the element of the Polypterids (single at the base, but immediately divaricating and with its limbs bordering an intervening cartilage which supports the pectoral and its basilar ossicles) must be homologous. But it is evident that the external elements of the so-called carpus of the teleosteoid Ganoids are homologous with that element in Polypterids. Therefore those elements cannot be carpal, but must represent the humerus. [Illustration: FIG. 75.--Shoulder-girdle of a Threadfin, _Polydactylus approximans_ (Lay and Bennett).] "The element with which the homologue of the humerus, in Polypterids, is articulated must be homologous with the analogous element in Dipnoans, and therefore with the _coracoid_. The coracoid of Polypterids is also evidently homologous with the corresponding element in the other Ganoids, and the latter consequently must be also _coracoid_. It is equally evident, after a detailed comparison, that the single coracoid element of the Ganoids represents the three elements developed in the generalized Teleosts (Cyprinids, etc.) in connection with the basis of the pectoral fin, and, such being the case, the nomenclature should correspond. Therefore the upper element may be named _hypercoracoid_; the lower, _hypocoracoid_; and the transverse or median, _mesocoracoid_.