ZOOLOGY AND BOTANY, MICKOSCOPY, ETC. 65 



forms the outer layer of the integument of the adult. Keller thinks 

 that the endoderm, in addition to forming part of the investment of 

 the ciliated chambers, as Ganin allows, forms the whole of the 

 flattened epithelial layer which forms the inner surface of the whole 

 of the gastric cavity. Both Keller and Ganin believe that the whole 

 of the mass of connective substance arises from the mesoderm. Led 

 by further observations to depart from the views which he once held, 

 Schulze now expresses his opinion that not only the collar-cells of 

 the so-called ciliated chambers, but the rest also of the flattened 

 flagellate epithelial cells are to be regarded as arising from the endo- 

 derm of the larva ; or, in other words, two different series of cells owe 

 their origin to the primitive endoderm, and the parts thus formed 

 clothe, in addition to the ciliated chambers, all the cavities, ducts, 

 and canals which arise from them, as far as the margin of the oscula. 

 So again we find that the layer of flattened epithelial cells, which 

 invests the free surface of the sponge (at any rate of P. monolopha), 

 and lines its clefts and canals, is ectodermal in origin ; it follows from 

 this that the ectodermal and endodermal layers meet one another at 

 two points, namely at the orifice of the ciliated chambers and around 

 the edges of the oscula. The intermediate tissue would seem to arise 

 from the mesoderm. 



The Plakinidte may be defined as Tetractinellidse with isolated 

 spicules (or with spicules not connected by a horny substance). The 

 spicules may be bi-, tri-, or quadri-radiate, but they would appear to 

 arise primitively from quadriradial spicules. Three genera are 

 included, PlaJcina, Flakortis, and PlaJcinastrella, with three (monolopha, 

 dilopha, trilopha), one (simplex), and one {copiosa) species respectively, 

 so far as is yet known. All these forms exhibit points of striking 

 resemblance in the histological details of their body-substance. The 

 great development of highly refractive granules in the region of the 

 ciliated chambers, in Plakortis and PlaMnastrella, are points in which 

 these two genera differ from PlaJcina, although there are indications 

 in this last genus of this structural character, and, indeed, after all, 

 we have rather to detect a gradual increase in the number of these 

 granules as we pass from PlaJcina monolopJia, through P. dilopJia and 

 P. trilopJia to PlaJcortis simplex. The author proceeds to point out the 

 relations exhibited by the spicules. 



Some points in the characters of the different species may be 

 noticed. P. monolopJia is of a pure white or delicate rosy colour, and 

 these colours appear to vary somewhat with the seasons. An ex- 

 tensive account is given of the development of this species. P. dilopha, 

 found at Trieste, has externally very much the same appearance as the 

 preceding form, but it is much smoother, and there is a striking differ- 

 ence in the characters of those spicules which belong to the class of 

 hard structures which Oscar Schmidt has distinguished as candelabra, 

 for two, and not one only, of the four primary rays are directed 

 towards the limiting surface, and both these divide into a number of 

 secondary rays. Similarly, P. trilopJia (Naples) has at least three 

 tufts of rays projecting outwards. In PlaJcortis no quadriradiate 

 spicules would appear to be present, and, as in PlaJcinastrella, the 



Ser. 2. —Vol. I. f 



