_Medusæ_, according to Weismann, they arise in the homologous layer, save where the medusoid form remains attached, and then they arise in the endoderm and migrate to the ectoderm: lack of specialization being in all cases implied. Then in the higher animals these same generative agents appear to be merely modified epithelium-cells--cells not remarkable for their complexity of structure but rather for their simplicity. If, by way of demurrer to this view, it be asked why other epithelium-cells do not exhibit like properties; there are two replies. The first is that other epithelium-cells are usually so far changed to fit them to their special functions that they are unfitted for assuming the reproductive function. The second is that in some cases, where they are but little specialized, they _do_ exhibit the like properties: not, indeed, by uniting with other cells to produce new germs but by producing new germs without such union. I learn from Dr. Hooker that the _Begonia phyllomaniaca_ habitually develops young plants from the scales of its stem and leaves--nay, that many young plants are developed by a single scale. The epidermal cells composing one of these scales swell, here and there, into large globular cells; form chlorophyll in their interiors; shoot out rudimentary axes; and then, by spontaneous constrictions, cut themselves off; drop to the ground; and grow into Begonias. Moreover, in a succulent English plant, the _Malaxis paludosa_, a like process occurs: the self-detached cells being, in this case, produced by the surfaces of the leaves.[30] Thus, there is no warrant for the assumption that sperm-cells and germ-cells possess powers fundamentally unlike those of other cells. The inference to which the facts point, is, that they differ from the rest mainly in not having undergone functional adaptations. They are cells which have departed but little from the original and most general type: such specializations as some of them exhibit in the shape of locomotive appliances, being interpretable as extrinsic modifications which have reference to nothing beyond certain mechanical requirements. Sundry facts tend likewise to show that there does not exist the profound distinction we are apt to assume between the male and female reproductive elements. In the common polype sperm-cells and germ-cells are developed in the same layer of indifferent tissue; and in _Tethya_, one of the sponges, Prof. Huxley has observed that they occur mingled together in the general parenchyma. The pollen-grains and embryo-cells of plants arise in adjacent parts of the meristematic tissue of the flower-bud; and from the description of a monstrosity in the Passion-flower, recently given by Mr. Salter to the Linnæan Society, it appears both that ovules may, in their general structure, graduate into anthers, and that they may produce pollen in their interiors. Moreover, among the lower _Algæ_, which show the beginning of sexual differentiation,