Now he found his theory to be in its pure form unworkable; he found, for example, that the skeleton of fishes could not be compared directly, bone for bone, with the skeleton of higher Vertebrates; he had to admit differences of position of whole sets of organs in the two groups, he had to admit various _metastases_, before he could bring the skeleton of fish into line. And these metastases are due to functional requirements--for example, the forward position of sternum and thoracic organs in fish is an adaptation to swimming. So he does not so much demonstrate the unity of plan of whole organisms as the unity of plan of particular corresponding parts of them. Thus he does not prove or attempt to prove that Articulates are in all points like Vertebrates, but simply that their skeleton is built upon the same plan as that of Vertebrates. The rest of the organs, while still comparable with the organs of Vertebrates, stand in different relations to the skeleton. An Articulate therefore, on his own showing, is not, _as a whole_, built upon the same general structural plan as a Vertebrate. Further, he does not always remain true to his principles, for he does not establish homologies of parts entirely by their connections but sometimes by their functions as well. Thus the sternum, or rather the complex of sternal elements, is defined and discovered in particular cases not by its connections only but also by its functions. The framework of the gills is homologised part by part with the framework of the lungs, not because the relations of the framework to the rest of the skeleton are the same in fish and air-breathing Vertebrates, but simply because gills are considered the equivalents of lungs--a comparison which is purely physiological. Even with these concessions to the functional view of living things, Geoffroy was unable to make good his contention that all animals are built upon the same plan. His arguments failed to carry conviction to his contemporaries, and Cuvier in particular subjected them to destructive, and indeed final, criticism. The paper, already referred to, in which Cuvier disposed of the transcendentalists' comparison of Cephalopods and Vertebrates is of great significance, for it states in the clearest way the radical opposition between the functional and the formal attitudes to living things.