§ 54c. As above implied, the individualities of the units are not wholly lost in the individuality of the aggregate, but continue, some of them, to be displayed in various degrees: the great majority of them losing their individualities more and more as the type of the aggregate becomes higher. In a slightly organized Metazoon like the sponge, the subordination is but small. Only those members of the aggregate which, flattened and united together, form the outer layer and those which become metamorphosed into spicules, have entirely lost their original activities. Of the rest nearly all, lining the channels which permeate the mass, and driving onwards the contained sea-water by the motions of their whip-like appendages, substantially retain their separate lives; and beyond these there exist in the gelatinous substance lying between the inner and outer layers, which is regarded as homologous with a mesoderm, amoeba-form protoplasts which move about from place to place. Relations between the aggregate and the units which are in this case permanent, are in other cases temporary: characterizing early stages of embryonic development. For example, drawings of Echinoderm larvæ at an early stage, show us the potential independence of all the cells forming the blastosphere; for in the course of further development some of these resume the primitive amoeboid state, migrate through the internal space, and presently unite to form certain parts of the growing structures. But with the progress of organization independence of this kind diminishes. Converse facts are presented after development has been completed; for with the commencement of reproduction we everywhere see more or less resumption of individual life among the units, or some of them. It is a trait of transitional types between _Protozoa_ and _Metazoa_ to lead an aggregate life as a plasmodium, and then for this to break up into its members, which for a time lead individual lives as generative agents; and sundry low kinds of plants possessing small amounts of structure, have generative elements--zoospores and spermatozoids--which show us a return to unit life. Nor, indeed, are we shown this only in the lowest plants; for it has recently been found that in certain of the higher plants--even in Phænogams--spermatozoids are produced. That is to say, the units resume active lives at places where the controlling influence of the aggregate is failing; for, as we shall hereafter see, places at which generation commences answer to this description.