In the shapes assumed by those rapidly-multiplying cells out of which each animal is developed, there is a conspicuous subordination to the surrounding actions. Fig. 294 represents the cellular embryonic mass that arises by repeated spontaneous fissions. In it we see how the cells, originally spherical, are changed by pressure against one another and against the limiting membrane; and how their likenesses and unlikenesses are determined by the likenesses and unlikenesses of the forces to which they are exposed. This fact may be thought scarcely worth pointing out. But it is worth pointing out, because what is here so obvious a consequence of mechanical actions, is in other cases a consequence of actions composite in their kinds and involved in their distribution. Just as the equalities and inequalities of dimensions among aggregated cells, are here caused by the equalities and inequalities among their mutual pressures in different directions; so, though less manifestly, the equalities and inequalities of dimensions among other aggregated cells, are caused by the equalities and inequalities of the osmotic, chemical, thermal, and other forces besides the mechanical, to which their different positions subject them. § 261. This we shall readily see on observing the ordinary structures of limiting membranes, internal and external. In Fig. 295, is shown a much-magnified section of a papilla from the gum. The cells of which it is composed originate in its deeper part; and are at first approximately spherical. Those of them which, as they develop, are thrust outwards by the new cells that continually take their places, have their shapes gradually changed. As they grow and successively advance to replace the superficial cells, when these exfoliate, they become exposed to forces which are more and more different in the direction of the surface from what they are in lateral directions; and their dimensions gradually assume corresponding differences. [Illustration: Fig. 295.] [Illustration: Fig. 296.] Another species of limiting membrane, called cylinder-epithelium, is represented in Fig. 296. Though its mode of development is such as to render the shapes of its cells quite unlike those of pavement-epithelium, as the above-described kind is sometimes called, its cells equally exemplify the same general truth. For the chief contrast which each of them presents, is the contrast between its dimension at right angles to the surface of the membrane, and its dimension parallel to that surface.