1. These experiments were conducted on Sycon raphanus at Naples, in order primarily to repeat H. V. Wilson’s work on coalescence and regeneration in monaxonid sponges, with a less specialised calcareous form. His method of,straining the chopped-up sponge through fine gauze was used, and by this means nothing came through except cells, singly or in twos and threes. These united together, mainly by means of actively amœboid cells, into small lumps of irregular shape, in which the various kinds of cells were confusedly mixed. This ended the process of reunion—one that does not occur normally in Nature. Next came that of reorganisation—similar in its main features to what occurs during reversal of the layers in the parenchymula larvæ of Calcarea ; the dermal cells migrated to the surface to form a flat epithelium round a solid central mass of quiescent collar-cells. The two layers of the sponge-body are now present in their definitive positions, and the subsequent period may be called one of redevelopment. The regenerate in this stage is very like a normal Sycon directly after metamorphosis, so that this redevelopment is practically the same as the normal post-larval development of the species. Spicules are formed, first monaxons, then triradiates ; a gastral cavity and finally an osculum appears. The only differences from normal development are that here spicules arise later (in the larva, he it noted, spicule-formation is certainly precocious), and that the regenerates, though making an attempt to fix, do not succeed in doing so permanently. None of the regenerates reached a heterocœle condition, but one lived and grew as a functioning sponge for several weeks. 2. By other methods, collar-cells were obtained nearly or quite pure. If large bits of unbroken gastral epithelium were taken, they bent back, and eventually rounded up to form perfect hollow spheres, with collars directed outwards. Similar spheres were also formed, but in a different way, when numerous small groups or single cells were taken. These united into solid lumps, which later swelled up to form spheres. Though some lived over a month, yet no other form of tissue was ever regenerated by them.