To begin with, it is not to a single dispersed particle or ultramicron that protoplasm has been likened, but to an emulsion, comprising both the dispersed particles and the dispersing medium, or, in other words, to the colloidal system as a whole. Moreover, even there the analogy is far from being perfect, and is confined exclusively, as Wilson has pointed out, to a rough similarity of structure and appearance. The colloidal system is obviously a mere _aggregate_ and not a _natural unit_ like the cell, and its dispersed particles (ultramicrons) do not multiply and perpetuate themselves by growth and division as do the living components or formed bodies of the cell. As for the single ultramicron or multimolecule of a colloidal solution, it may, indeed, be a natural unit, but it only resembles the cell in the sense that, like the latter, it is a complex of constituent molecules. Here, however, all resemblance ceases; for the ultramicron does not display the typically vital power of self-perpetuation by growth and division, which, as we have seen, is characteristic not only of the cell as a whole, but of its single components or organelles. Certainly, the distinctive phenomena of colloidal systems cannot be interpreted as processes of multiplication. There is nothing suggestive of this vital phenomenon in the reversal of phase, which is caused by the addition of electrolytes to oil emulsions, or in gelation, which is caused by a change of temperature in certain hydrophilic colloids. Thus the addition of the salt of a bivalent cation (_e.g._ CaCl₂ or BaCl₂) to an oil-in-water emulsion (if soap is used as the emulsifier) will cause the external or continuous phase (water) to become the internal or discontinuous phase. Vice versa, a water-in-oil emulsion can be reversed into an oil-in-water emulsion, under the same conditions, by the addition of the salt of a monovalent cation (_e. g._ NaOH). Solutions of hydrophilic colloids, like gelatine or agar-agar, can be made to “set” from the semifluid state of a hydrosol into the semisolid state of a hydrogel, by lowering the temperature, after which the opposite effect can be brought about by again raising the temperature. In white of egg, however, once gelation has taken place, through the agency of heat, it is impossible to reconvert the “gel” into a “sol” (solution). In such phenomena, it is, perhaps, possible to see a certain parallelism with some processes taking place in the cell, _e. g._ the osmotic processes of absorption and excretion, but to construe them as evidence of propagation by growth and division would be preposterous.