For the cocoons are rather bulky for this kind of porterage. Moreover they are apt to be deposited rather deep down and among the roots of grasses, and in situations where they are not so likely to become entangled in the feet of drinking birds. Assuming, however, that these difficulties can be got over there remains another difficulty. A single cocoon among the terrestrial Oligochaeta does not contain a large number of embryos, as has been pointed out on a previous page. It is true that _Allolobophora foetida_ has six within one cocoon, but most of our indigenous forms have but from one to three embryos in a single cocoon. Thus, if successfully imported, it is hardly likely that the developed embryos scattered after their emergence would come together for breeding purposes; and in cocoons with but one embryo the accidental importation in this way would have to be very frequent to produce any result. The case here is exactly the reverse of that afforded by the aquatic families (or many of them). In these Annelids the attachment of the cocoon to water plants, which are liable to be entangled in the feet of shore-frequenting birds, would tend to favour migration. And in addition to this the cocoons are naturally smaller and often contain a considerable number of embryos. We are to note that the aquatic forms are on the whole distinctly wider in their range than are the earthworms. CHAPTER IX THE GEOGRAPHICAL DISTRIBUTION OF EARTHWORMS The facts referred to and considered in the last chapter lead to further observations upon the geographical distribution of this group of animals and suggest problems for solution. It is not the place here to give a general sketch of the division of Biology termed Zoogeography; but a few general conclusions must be laid before the reader in order to render what follows intelligible. It is universally agreed that the range in space (and in time also) of a given species of animal (or plant) is as much a part of its scientific definition as are its anatomical characters. A description for instance of _Acanthodrilus ungulatus_ is incomplete without a reference to the fact that it occurs in, and is confined to, the island of New Caledonia.