336 On the Development of the Marine Sponges. 



embryo came from, viz. Halichondria simulans, under the 

 names (as before stated) of Chalina simulans in one part, and 

 Isodictya simulans in another (B. S. I. c. plate xix. fig. 299). 



At this point of development the spicules, now in a naked 

 form, project a little beyond the " investing membrane " or 

 ectodermal layer (fig. 32, c c) ; and could I have brought a 

 4^-inch compound power to bear upon it in this development, 

 as in the development of Spongilla from the seed-like body, in 

 a watch-glass, I might have observed the pores, and finally 

 the composition of the investing membrane or ectodermal 

 layer itself, which again, strangely, in Spongilla I have de- 

 scribed as consisting of flat cells like Amcebce, forming a 

 fl foliated arrangement not unlike a compressed layer of multi- 

 fidous leaves ever moving and changing their shapes" (?.c.p. 25, 

 pi. i. fig. 7). I say " strangely," because this membrane, in 

 the embryo of Halichondria simidans, must also be composed 

 of polymorphic cells, although with their cilia now retracted. 



Of the same nature, composition, and functions as the invest- 

 ing membrane in Spongilla we may, then, fairly assume that in 

 the embryonal sponge of Halichondria simidans to be ; and I 

 would recommend any one who wishes to make himself con- 

 fident of this to consult my paper and illustrations on the 

 ultimate structure of Sp>ongilla (op. et loc. cit.). 



The single vent, now the end of the branched excretory 

 canal-system, may be observed to traverse the cavity of the 

 investing membrane and to open on the surface (tig. 33, a). 



Thus the passing of the embryo into the perfected sponge 

 may be considered to have become complete. 



On the 11th of August I sought, as before, eagerly for a 

 current of particles issuing from the vent, but could observe 

 none ; and fancying that the development had become sta- 

 tionary (in fact, that a slight retraction all round the young 

 sponge indicated approaching dissolution), I determined not 

 to try to feed it with indigo in order that I might see if it 

 already contained fully-formed spongozoa and ampullaceous 

 sacs, but at once tore it to pieces on a slide in sea-water for 

 this purpose, well knowing that a little delay in doing this 

 would expose the whole structure to the ruinous influence of 

 hosts of devouring animalcules. 



This was done, and then the structure of the parent sponge 

 was observed to have become fully developed (fig. 34). The 

 skeleton structure was found to consist of a reticulated horny 

 investment supporting bundles of the acerate spicule of the 

 species (fig. 34, a a), some of which, as before mentioned, 

 projected beyond the confines of the opaque or parenchyma- 

 tous portion of the sponge so as to support the investing 



