and tlieir Relatioiishq) to the Corals. Ill 



The mode of production of the spores or so-called ova of the 

 sponges has hitherto been unknown. In my monograph I 

 shall demonstrate that thej proceed directly from the flagellate 

 cells, and consequently are products of diferentiation of the 

 entoderm^ or metamorphosed fagellate cells. The simple and 

 extremely significant fact that the reproductive cells are pro- 

 duced, by division of labour, from the nutrient vibratile cells 

 of the entoderm or vegetative germ-lamella applies also to 

 the sponges equally with the Acalephs. According to Kol- 

 liker, the spores of Dunstervillia and Tarrus lie outside the 

 vibratile epithelium in the ectoderm. But they only get there 

 when, from the increase of their bulk, they can no longer find 

 room among the surrounding flagellate cells of the entoderm. 

 They then project sometimes into the ectoderm and sometimes 

 into the lumen of the canals. I have never found special 

 spore-capsules in the Calcispongiaj, but the spores may deve- 

 lope themselves from the flagellate cells on the most difl^erent 

 spots in the entoderm. What Lieberklihn describes in St/cum 

 as a special " receptacle for the ova, without demonstrable 

 structure," I have never seen, and I suppose that these asserted 

 spore-capsules are ti-ansversely cut canals. 



As Kolliker has already pointed out, the spores of the 

 sponges have a remarkable resemblance to large ganglionic 

 cells. This is due to the fact that the protoplasm of the cells 

 emits from the periphery polymorphic branched processes. 

 The spores of the Calcispongue resetahle large Amoebae, and 

 jjerform anioehoid movements^ by extending and retracting 

 such branched processes. In a state of repose, they are sphe- 

 rical or polyhedral. Each spore possesses a very large, usu- 

 ally spherical, and limpid nucleus. This encloses a large, 

 round, dark, nucleolus, and this, again, a distinct nucleolinus. 



The Spongice are in jxirt s2)orip)arous and in part viviparous. 

 In the sporiparous sponges {e. g. Leucosoleniaj Clistolynthus) 

 the mature spores drop out of the entoderm into the stomachal 

 cavity or into the parietal canals issuing from the latter, and 

 are then cast forth through the mouth in the forms which are 

 provided with a mouth, whilst in astomatous sponges they 

 creep out through the cutaneous pores. In the latter case their 

 amoiboid movements will be of essential assistance to them. 



In the viviparous sponges (e. g. Olgnthus^ Clathrina) a 

 spherical body (embryo), composed entirely of similar naked 

 nucleated cells, is produced from the simple spore-cell by con- 

 tinued division (" segmentation ") within the body of the 

 sponge (either in the stomach or in the parietal canals issuing 

 from it. Each of the cells situated on its surface emits a fila- 

 mentous process, and thus becomes a flagellate cell. Then 



