In the only illustration hitherto adduced--viz. that of deers' horns--the series of changes from a one-pronged horn to a fully developed arborescent antler, is a series which takes place during the adult life of the animal; for it is only when the breeding age has been attained that horns are required to appear. But seeing that every animal passes through most of the phases of its development, not only before the breeding age has been attained, but even before the time of its own birth, clearly the largest field for the study of individual development is furnished by embryology. For instance, there is a salamander which differs from most other salamanders in being exclusively terrestrial in its habits. Now, the young of this salamander before their birth are found to be furnished with gills, which, however, they are never destined to use. Yet these gills are so perfectly formed, that if the young salamanders be removed from the body of their mother shortly before birth, and be then immediately placed in water, the little animals show themselves quite capable of aquatic respiration, and will merrily swim about in a medium which would quickly drown their own parent. Here, then, we have both morphological and physiological evidence pointing to the possession of gills by the ancestors of the land salamander. It would be easy to devote the whole of the present chapter to an enumeration of special instances of the kinds thus chosen for purposes of illustration; but as it is desirable to take a deeper, and therefore a more general view of the whole subject, I will begin at the foundation, and gradually work up from the earliest stages of development to the latest. Before starting, however, I ask the reader to bear in mind one consideration, which must reasonably prevent our anticipating that in _every case_ the life-history of an individual organism should present a _full_ recapitulation of the life-history of its ancestral line of species. Supposing the theory of evolution to be true, it must follow that in many cases it would have been more or less disadvantageous to a developing type that it should have been obliged to reproduce in its individual representatives all the phases of development previously undergone by its ancestry--even within the limits of the same family. We can easily understand, for example, that the waste of material required for building up the useless gills of the embryonic salamanders is a waste which, sooner or later, is likely to be done away with; so that the fact of its occurring at all is in itself enough to show that the change from aquatic to terrestrial habits on the part of this species must have been one of comparatively recent occurrence. Now, in as far as it is detrimental to a developing type that it should pass through any particular ancestral phases of development, we may be sure that natural selection--or whatever other adjustive causes we may suppose to have been at work in the adaptation