When one liquid, say spirit, is poured into another, say water, the two gradually mix. If we suppose these liquids to consist of a number of molecules, then, mixture may be taken to be the result of the displacement say of one molecule of water by one molecule of spirit, and so, throughout the whole quantity of liquid, there is displacement and replacement till at length equilibrium is restored and a thorough diffusion results. This power of diffusion does not always exist. The molecules of water and of oil will not mix or diffuse freely through each other. Water containing carbonic acid gas will not mix, in this sense of the term, with water containing acetate of lead. It may be a truism to say, that for the process of diffusion the liquids must be diffusible, but the fact must be carefully borne in mind in all questions relating to the feeding of plants. In the case of plants, the phenomenon of diffusion, or the gradual admixture of two liquids of different natures, is complicated by the presence of a membrane in the shape of the cell-wall. The water from the outside has to pass through the membrane to reach the protoplasm on the other side. Speaking broadly, there are no holes in the membrane through which the water can pass. Ingress is secured by that process of diffusion to which reference has just been made, and by virtue of which the molecules of the membrane and the molecules of the water shift and change places; the space that was occupied by a molecule of membrane is now occupied by a molecule of water, and _vice versa_. The access, therefore, of water into the interior of a closed cell is the result of the process of diffusion. Where two liquids mix without any intervening membrane, the mixture is called diffusion simply; where there is an intervening membrane, the diffusion process is known as “osmosis.” The raw material (the term is not quite accurate, but for illustration sake it may pass) is that very marvelous substance now called “protoplasm.” We must leave it to chemists and microscopists to explain its composition and indicate its appearance. Diffusion is not equal or alike in all cases; it depends upon the extent to which the two liquids are diffusible, upon their different densities, upon temperature, and a variety of other conditions. So, in the case of osmosis, we have not only the nature of the two fluids to consider, but their relation to the membrane that separates them. The membrane may be much more permeable to one of the two fluids than to the other. Thus, in the case of a living cell, the membrane or wall is much more permeable to water than it is to protoplasm; and so it happens that, while water readily penetrates the membrane and diffuses itself in the protoplasm, protoplasm does not nearly so readily permeate the membrane as the water. Ingress of water is easy and of constant occurrence, egress of protoplasm is rare and exceptional.