Secondary aggregates in which the lives of the units are more subordinate to the life of the whole, carry this distinction further. The leading physiological trait of every cœlenterate animal is the divisibility of its substance into endoderm and ectoderm--the part next the food and the part next the environment. Fig. 147 (§ 201), representing a portion of the body-wall of a _Hydra_ seen in section, gives some idea of this fundamental differentiation. The creature consists of a simple sac, the cavity of which is in communication with the surrounding water; and hence the unlikeness between the outer and inner layers has not become great. The essential contrast is that between the differentiated parts of what was originally the same part--a uniform membrane composed of juxtaposed cells. For here, indeed, we are shown unmistakably how the primary contrast of structures follows upon the primary contrast of conditions. The ordinary form from which low types of the _Metazoa_ set out, is a hollow sphere formed of cells packed side by side--a blastula, as it is called: all these cells being similarly exposed to the environment. The blastula presently changes into what is called a gastrula--a form resulting from the introversion of one of the sides of the blastula. If there be taken a small ball of vulcanized india-rubber, say an inch or more in diameter, and having a hole in it through which the air may escape, and if one side of it be thrust inwards so as to produce a cup, and if the wide opening of the cup be supposed to contract, thus becoming a narrow opening, there will result something like the gastrula form. Manifestly that part of the original layer which has become internal is differently conditioned from the rest which remains external: the one continuing to hold converse with the forces of the environment, while the other begins to hold converse with the nutritive matters taken into the sac-formed chamber--the archenteron or primitive stomach. Interesting evidence of the primitive externality of the digestive cavity is yielded by the fact that whereas the blastula consisted of ciliated cells, and whereas the ciliation persists throughout life on the outer layer, or parts of it, in sundry low types--even in some Chætopods--it persists also on the alimentary tract of sundry low types: not only in the _Hydra_ but commonly in Nemertines, in some _Platyhelminthes_, and even in some leeches. Besides being enabled thus to understand how an aggregate of _Amœba_-form units, originally consisting of a single layer, may pass into an aggregate consisting of a double layer; we may also understand under what influences the transition takes place. If the habit which some of the primary aggregates have, of wrapping themselves round masses of nutriment, is followed by a secondary aggregate, there will naturally arise just that re-differentiation which the _Hydra_ shows us.