710 SUMMARY OF CURRENT RESEARCHES RELATING TO 



imnibcr of osciila, wliich are found on both sides of tlio sponge, and 

 which vary a good deal in size. No afferent pores couhl be detected 

 on the surface of tlie sponge. There are various forms of siliceous 

 structures, into the detailed account of whicli the author enters very 

 fully. Ho discusses also their mode of growth, and comes to the 

 conclusion that the skeletal spicules probably grow by apposition, 

 while those of the cortex are not essentially increased in size by such 

 process. 



Prof. Noll was unable to make out the ectoderm, but he ascribes this 

 to the mode of preparation, as he does not accept the doctrine of Gotto 

 that the ectoderm of all sponges is lost during mctamori)hosis, and ho 

 brings evidence afforded by his own observations on Sjiongilla jlnviatilis 

 as opposing it. The surface of the Dcsmacidun is thin and traii8])arent ; 

 it generally lies close to the parencliyma, and cannot be easily torn off 

 in large shreds. Sometimes the layer ai)i)ears to be merely formed of a 

 homogeneous ground substance with a few cell-nuclei, but in other cases 

 there are cellular elements, some contractile fibres, or non-nucleated 

 fibres. "Where the clear ground-substance is predominant, numerous 

 cell-nuclei of various sizes are imbedded in it ; these do not colour 

 strongly, and never lie so close to one another that their boundaries 

 touch. Tlie non-uucloated fibres generally lie close to one another and 

 form bands which run in various directions. The contractile fibres 

 appear to be the elongated terminal jioles of long spindle-shaped cells ; 

 the nuclei of these cells are oval, and the cell-contents finely granular. 

 One would be inclined to sjieak of these as muscle-cells, if they could be 

 shown to be provided with nerves. In any case it is possible that they 

 have some reflex activity, and changes in the form of the surface of the 

 sj)onge may be often observed. 



Around the osculum there is a circlet of the so-called muscle-cells, 

 and by their elongation the orifice could certainly be narrowed. The 

 contractile fibres are wanting in the neighbourhood of the smaller 

 openings which serve as incurrent orifices. 



The parenchyma is very well devcloi)ed, and forms the chief part of 

 the sponge ; in it the cellular elements are pretlominant, and the ground- 

 substance is considerably reduced ; compared with those of Spoiujilla, 

 the cells are proportionately small, but they vary considerably in form 

 and size. Non-nucleated protoplasmic corpuscles ii.ake up the chief part 

 of the parenchyma ; between them a number of free cell-nuclei are to 

 be seen in the parenchyma, and these are all spherical in form. Complete 

 cells with protoi)lasm and nucleus, but in all cases without a membrane, 

 are not so numerous as the bodies just mentioned. Occasionally there 

 are two nuclei in one cell ; wandering cells are also to be seen, and 

 they may or may not have a nucleus; indeed, it might almost be thought 

 tbat the nuclei and protoplasm of the cells can lead an iudepemlent life. 



With regard to the formation of sjiicules, the author concludes that 

 definite cells are set apart for the purpose ; these silicoblasts elongate, 

 their contents clear up, and the central filament first appears. As the 

 delicate membrane of the body of a RhizojTOd conditions the form of the 

 calcareous shell, and as the test of the diatom is preformed by the 

 delicate cell-meudjrane which serves as its basis, so we may, with 

 Bowerbank, call the central filament a membrane formed internally by 

 the cell. Its form depends on that of the silicoblast, and so we see the 

 spicular mother-cells of the spicula of SponrjUla elongate like the 



