This prolongation of cells into long cylinders (called fibres) is much more remarkable in the crystalline lens. The fibres or cylindrical cells of the lens, however, themselves undergo very important modifications, inasmuch as they often become flattened on two sides into bands, and then the margins of these bands become denticulated. This serration is probably produced by a more forcible expansion, and therefore bulging-out of the walls of these bands at different points, which follow each other at pretty regular distances, whilst the intervening points, situated close to them, remain stationary. All the different stages of this serration, may be observed in the lens of the fish, if the fibres are examined from the exterior towards the centre of the structure. Now, in this flat and serrated condition, the cells of the crystalline lens perfectly resemble those of the epidermis of some grasses, and this accordance with indubitable vegetable cells is a proof that, despite the modifications which they undergo, they do not lose their cellular character. If the explanation I have given of the mode in which the serration is produced be correct, it will not materially differ in principle from the elongation of the cells into cylinders and fibres. For, in the latter case, a more forcible expansion of the cells is likewise presumed to take place in certain situations: the sole difference being, that in the latter case it takes place only at one or two opposite points of a cell, whereas with the serration it occurs at many separate ones. At this stage of our inquiry, we are reminded of the form of many pigment-cells, in which this expansion of the cell, at certain spots, takes place on several sides, and in a far higher degree, causing the cell to assume an irregular stellated form. The prolongations of these cells, however, retain their character as hollow processes, even when almost as minute as the fibres of cellular (areolar) tissue.
The distinction between cell-membrane and cell-contents is nowhere more distinctly defined than in the fully-developed cells of this class. In the perfected cells of the pith of feathers, for example, it is as marked as we ever find it to be in plants. When traced backwards to their earliest stages of development, their true cellular formation scarcely admits of a doubt, although the cell-membrane, for reasons given at page 36, cannot be so clearly distinguished. The elementary
