The obvious question, Why do some Crustacea pass through a complicated metamorphosis while others do not? is, like many obvious and simple questions, one of the most difficult to answer. It will be pointed out later, in dealing with the fresh-water Crustacea, that one of the most general characters of fresh-water animals as compared with their marine allies is the absence of free-swimming larval stages. This applies, for instance, to the case of the Crayfishes and the marine Lobsters, and to that of the River Crabs, as compared with those which live in the sea. But it does not apply to all fresh-water Crustacea, and, on the other hand, there are many cases of direct development in marine species. Some of the advantages gained by the possession of free-swimming larval stages are obvious enough. Many Crustacea which live on the sea-bottom, and are not very powerful swimmers, have their progeny scattered far and wide by winds and currents while in the surface-living larval stages. In the extreme case of the Barnacles, which are fixed to one spot when adult, a locomotive larval stage is clearly a necessity. But, here as elsewhere, to demonstrate the usefulness of any character is to go only a very little way towards explaining its origin. Moreover, the mere necessity for a locomotive larva throws no light on the remarkable resemblances between the larval stages of widely different species. In the adult state, a Branchiopod, a Copepod, an Ostracod, a Barnacle, and a Penæid Prawn, are separated by enormous differences of form and structure; yet, as we have seen, all these are hatched from the egg as six-limbed nauplius larvæ differing from each other only in trivial details. It seems hardly possible to imagine any other interpretation of this very striking fact than is afforded by the theory of Evolution. We are forced to assume that all these diverse forms of Crustacea are descended from very similar or identical ancestral types, and that the modifications arising in the course of their evolution have affected the adult but not the larval stages. Some naturalists would go farther than this, and would apply the so-called "theory of recapitulation" to the larval stages of the Crustacea. According to this theory, the stages in the development of any animal tend to recapitulate, more or less closely, the history of the race. Thus it is assumed, for instance, that the nauplius reproduces the structure of a six-limbed ancestral form, from which, in the distant past, all the diverse branches of the Crustacean class took their origin. There are, however, considerable difficulties in the way of this view. That some such ancestral type did exist may be regarded as tolerably certain; that it resembled in its adult state the nauplius larvæ of present-day Crustacea is, on the whole, unlikely; but it is not at all improbable, whatever its adult structure may have been, that it hatched from the egg as a nauplius larva.