elements, the sternum and the ribs, must have moved with them, and are hence to be found in front of the pectoral girdle. [Illustration: FIG. 1.--Hyoid Arch of the Conger. (Original.)] Geoffroy's next step is to point out that the only possible homologues of sternal ribs are the branchiostegal rays, which arise from the large bones of the hyoid arch. If these are sternal ribs, the bones to which they are attached must be the hyo- and hyposternals or "annexes," the bones from which in birds the ribs take their origin. The unpaired sternal bone (_urohyal_) cannot be homologous with the entosternal, for it has no connections with the annexes. He decides that it must represent the episternals, for in some young birds there is a two-headed episternal to which two strong tendons are attached, just in the same way as the unpaired piece in fish is bound to the bones of the hyoid by two tendons. "Thus it is not the sternum as a whole that has shifted in front of the clavicles and covered with its side pieces the gills placed there; it is a piece exclusively piscine, in the sense that it is only in the class of fishes that it reaches the _maximum_ of its development" (p. 83). To sum up, the sternum in all four vertebrate classes is composed of the same elements, arranged always in the same way. "One is ... led to the conception of an ideal type of sternum for all Vertebrates, which then, considered from a lower standpoint, resolves itself into several secondary forms according as the whole or the majority of the constituent materials are employed, or even as these elements come to change their respective dimensions or proportions" (p. 134). As to the elementary constituents, "they give proof of individuality, and sometimes even, in certain abnormalities, of independence, and rise to the level of primary organisatory materials" (p. 132). What holds good for the sternum holds good for other organs--and accordingly the unity of plan and composition can be demonstrated for all the organs of Vertebrates. Soon after the publication of the _Philosophie anatomique_ (1818) Geoffroy went further in his search for unity, and maintained that the structure of insects and Crustacea could be reduced to the vertebrate type. He proposed to replace Cuvier's classification of the animal kingdom into the four large groups, Vertebrata, Mollusca, Articulata, and Radiata by the following classification:--[90] Hauts-Vertebres (Vertebrata, Cuv.). Vertebres / \ Dermo-Vertebres (Articulata, Cuv.). Mollusques (Mollusca, Cuv.). Invertebres / \ Rayonnes (Radiata, Cuv.).