and their Relationship to the Corals. 109 



notwithstanding their further difterentiation to various higher 

 structures. 



In all Calcispongiffi without exception (although in some 

 more distinctly than in others), the I'undamental and original 

 difference of the two formative membranes stands out so dis- 

 tinctly, and may be so readily and clearly traced in their fur- 

 ther divergence, even up to the most highly developed forms, 

 that it may be at all times visibly demonstrated. Consequently 

 it has not escaped those naturalists who have most carefully 

 investigated the structure of the Calcispongias. Here and there 

 they all speak of the different layers of the body-wall ; but 

 none of them has indicated their general and genetic signifi- 

 cance, and no one has perceived that the entoderm produces 

 exclusively the epithelium of the canal-system, which per- 

 forms the function of nutrition, and the cells serving for re- 

 production, and the ectoderm all the other cells. For this 

 reason I may be permitted here to adduce some special cir- 

 cumstances connected Avith the structure of the body in the 

 Calcispongiaj, the detailed description of which, and their elu- 

 cidation by figures, I reserve for my monograph. 



The entoderm^ or inner formative membrane of the Calci- 

 spongiaj, produced from the inner cell-layer or vegetative 

 germ-lamella of the embryo, originally lines the whole inner 

 surface of the nutrient canal-system or gastrovascular system 

 in the form of a single continuous cell-layer of flagellated epi- 

 thelium. By the expression flagellated epithelium (Geissel- 

 Epithel, e^jithelium flagellatum) I understand an epithelial 

 cell-layer, each cell of which hears a single vihratile hair (fla- 

 gellum), in contradistinction to ciliated epithelium (Wimper- 

 Epithel, epithelium ciliatum), each cell of which bears two or 

 more vihratile hairs (Wimpern, cilia). Flagellated and ciliated 

 epithelia are to be distinguished as two different modifications 

 of vihratile epithelium (Flimmer-Epithel, epithelium vibrato- 

 rium). In all sponges the vihratile epithelium appears to occur 

 exclusively in the form, of flagellated epithelium^ and never in 

 that of ciliated epithelium. This applies both to the vihratile 

 cells which line the inner surface of the canal-system and to 

 those which clothe the outer surface of the vihratile swimming 

 larva. In both cases the epithelial cells are always mono- 

 trichal, flagellate cells, and never polytrichal, ciliate cells. The 

 flagellate cells of the sponges are perfectly naked and mem- 

 braneless ; their protoplasm passes directly into the long fla- 

 gellum, which is thicker at the base. In the flagellate cells 

 I have never failed to find a distinct nucleus. It is usually of 

 very considerable size, one-half or two-thirds as large as the 

 cell. Generally the flagellate cells line the walls of the canal- 

 system only in a single layer ; rarely several layers are super- 



