The description of the movements of the tentacles was interrupted in the third chapter for the sake of describing the process of aggregation. This process always commences in the cells of the glands, the contents of which first become cloudy; and this has been observed within 10 s. after a gland has been excited. Granules just resolvable under a very high power soon appear, sometimes within a minute, in the cells beneath the glands; and these then aggregate into minute spheres. The process afterwards travels down the tentacles, being arrested for a short time at each transverse partition. The small spheres coalesce into larger spheres, or into oval, club-headed, thread- or necklace-like, or otherwise shaped masses of protoplasm, which, suspended in almost colourless fluid, exhibit incessant spontaneous changes of form. These frequently coalesce and again separate. If a gland has been powerfully excited, all the cells down to the base of the tentacle are affected. In cells, especially if filled with dark red fluid, the first step in the [page 265] process often is the formation of a dark red, bag-like mass of protoplasm, which afterwards divides and undergoes the usual repeated changes of form. Before any aggregation has been excited, a sheet of colourless protoplasm, including granules (the primordial utricle of Mohl), flows round the walls of the cells; and this becomes more distinct after the contents have been partially aggregated into spheres or bag-like masses. But after a time the granules are drawn towards the central masses and unite with them; and then the circulating sheet can no longer be distinguished, but there is still a current of transparent fluid within the cells. Aggregation is excited by almost all the stimulants which induce movement; such as the glands being touched two or three times, the pressure of minute inorganic particles, the absorption of various fluids, even long immersion in distilled water, exosmose, and heat. Of the many stimulants tried, carbonate of ammonia is the most energetic and acts the quickest: a dose of 1/134400 of a grain (.00048 mg.) given to a single gland suffices to cause in one hour well-marked aggregation in the upper cells of the tentacle. The process goes on only as long as the protoplasm is in a living, vigorous, and oxygenated condition.