320 THE STRUCTURE OF EVOLUTIONARY THEORY
4 - 14. The key plate from Owen's 1849 monograph on the nature of limbs. The archetype, built of
a series of vertebrae, shown at the top, with skeletons of a fish and a reptile below. In his boldest
move, Owen attempts to derive the entire limb of later vertebrates from the single diverging ray
of each vertebral element in the archetypal form. The diverging ray is the little spike projecting
upward at about 10 o'clock from the junction of two vertical elements below each vertebral
centrum of the archetype. (Author's collection.)progressively superadded structures and perfections in higher reptiles and in
mammals" (p. 70).
But these strained homologies then incur a second, equally serious problem in
requiring a pronounced shift of position among vertebrate classes—an
interpretation inconsistent with the formalist principle that topology and
connection serve as the primary criteria of homology. Geoffroy had developed his
concept of metastasis (see p. 300) to explain exceptions in the same troubling
example, and Owen followed this continental solution. The pectoral girdle of fishes
attaches to the rear of the skull. In fact, Owen regarded the bones of the girdle as
the haemal portions of the fourth, or occipital, skull vertebra. (The arm and hand
arise from the diverging ray of this vertebra and also become parts of the head by
homology.) Owen recognizes the counterintuitive oddity of such a claim, but must
follow the formalist logic: "However strange and paradoxical the proposition may
sound, the scapular arch