We have seen what were von Baer's criticisms of the theory of parallelism. If we turn to the later writings of Cuvier we find the essential criticism expressed in similar terms. Speaking of an attempt which had been made to show that fish were molluscs developed to a higher degree, he wrote in 1828,[179] "Let us draw the conclusion that even if these animals can be spoken of as ennobled molluscs, as molluscs raised to a higher power, or if they are embryos of reptiles, the beginnings of reptiles, this can be true of them only in an abstract and metaphysical sense, and that even this abstract statement would be very far from giving an accurate idea of their organisation." From the fact that the respiratory and circulatory organs of fish greatly resemble those of tadpoles the conclusion has been drawn that fish are in a sense embryos of Amphibia (p. 547). But this manner of viewing things is none the less vicious, "for this reason ... that it considers only one or two points and neglects all the others" (p. 548), and is directly contrary to common sense. There is never a recapitulation of total organisations, only at the most of single organs. It will be remembered that Cuvier opposed and demolished the theory of the _Echelle des etres_, not only by showing that there were in Nature four entirely different plans of animal structure, but also by demonstrating that even the animals of each single _Embranchement_ could not readily be arranged in one series, that a serial arrangement was really valid only for their separate organs. Von Baer also held that there are four distinct types of structure; he, too, combated the idea of gradation within the limits of the type. In so far as species represent successive stages in the development, the _Ausbildung_, of the type, so far can the idea of a scale of beings be applied. But the members of a type follow not one line of evolution but several diverging lines, in direct adaptation to different environmental conditions, so that a serial arrangement of them is not as a rule possible. It may be possible to establish a serial arrangement of single organs from the simplest to the most complex. But each organ or organ-system will require a different serial arrangement, for the different systems vary on different lines and an animal may be highly developed in respect of one system and little developed in respect of all the others. Man, for instance, is the highest animal only in respect of his nervous system. The idea of the scale of beings has therefore only a very limited application even within the limits of the type. Applied to the whole animal kingdom it becomes merely absurd.