He recognised, however, that the idea of a simple scale of beings is only an abstraction, and that the true repetition is of organs rather than of organisms. He was willing even to admit, at least in the later pages of his memoir, that there might be not one animal series but several parallel series, as had been suggested by Isidore Geoffroy St Hilaire (p. 749). In general, his views are now less dogmatic than they were in his earlier writings, but they are not for all that changed in any essential. For, in summing up his main results, he writes, "The whole animal kingdom can in some measure be regarded ideally as a single animal, which, in the course of formation and metamorphosis in its diverse manifestations, here and there arrests its own development, and thus determines at each point of interruption, by the very state it has reached, the distinctive characters of the phyla, the classes, families, genera, and species" (p. 833).[323] To settle the dispute pending between two of its most illustrious members, the Academy proposed in 1853, as the subject of one of its prizes, "the positive determination of the resemblances and differences in the comparative development of Vertebrates and Invertebrates." A memoir was presented the next year by Lereboullet[324] which met with the approval of the Academy in so far as its statements of fact were concerned, but seemed to them to require amplification in its theoretical part. But even in this memoir Lereboullet was able to show that the balance of evidence was greatly in favour of Milne-Edwards' views, and his general conclusions in 1854 were that "in the presence of such fundamental differences, one is obliged to give up the idea of one single plan in the formation of animals; while, on the contrary, the existence of diverse plans or types is clearly demonstrated by all the facts" (p. 79). To fulfil the Academy's requirements, Lereboullet continued his work, and in 1861-63 he published a series of elaborate monographs[325] on the embryology of the trout, the lizard and the pond-snail _Lymnaea_, and rounded off his work with a full discussion[326] of the theoretical questions involved. In this considered and authoritative judgment he completely disposed of Serres' theories of the unity of plan and the unity of genetic formation. Except in the very earliest stages of oogenesis there is no real similarity between the development of a Zoophyte, a Mollusc, an Articulate and a Vertebrate, but each is stamped from the beginning with the characteristics of its type. The lower animals are not, and cannot possibly be the permanent embryos of the higher animals. "The results which I have obtained," he writes, "are diametrically opposed to the theory of the zoological series constituted by stages of increasing perfection, a theory which tries to demonstrate in the embryonic phases of the higher animals a repetition of the forms which characterise the lower animals, and which