mVERTEBBATA, CRTPTOGAMIA, MICROSCOPY, ETC. 105 



ectoderm and one the endoderm. The endodermal cell becomes en- 

 capsuled by the multiplication of the ectoderm cells round it, and 

 then proceeds to increase by fission, forming a single row of cells, 

 and then — the ectoderm ceasing to multiply its cells — a double row, 

 and so on, until the whole mass becomes globular from the internal 

 development of endoderm cells. Now the ectoderm resumes its 

 activity and becomes bi-laminate, the outer layer constituting the 

 small-stellate-spicule layer, the inner forming the source of the futui-e 

 large-stellate- and fibre-layers, and of the medullary tissue. 



The gemmae, thus originated, on reaching a diameter of 1 mm. are 

 supported by a pedicle of spicules ; they then consist of an epithelium, 

 a layer of small stellates, below this the fibre- and large-stellate 

 cells, and innermost the medullary tissue, the whole diameter being 

 traversed radially by linear spicules. 



The important modifications occurring in the next — the 2 mm. 

 diameter— stage are the development of the water-vascular system and 

 the fuller differentiation of the spicular and cellular elements. 



The bud reaches the exterior by its contact with a bundle of the 

 radiating spicules ; these gradually move towards the perii)hery, 

 being met by an invagination of the more superficial small-stellate 

 layer; the gemma arrives opposite to this opening, and is thus 

 extruded by its spicule-bundle from the sponge. This method of 

 development of buds is especially interesting from the few exact 

 parallels with which it meets in other animals, the case of Loxosoma 

 as lately given by Professor 0. Schmidt * being one of these. 



This method of development by germ-layers in a hud is probably 

 more ancient than the same process exhibited by an ovum. The 

 mesoderm, according to the results here set forth, is double, which 

 F. E. Schulzet has hitherto not admitted, although he is dis- 

 tinguished among the upholders of the three-layered structure of 

 sponges. Certain varieties of T. h/ncurium, viz. var. vUlosa Schmidt, 

 MS., and var. Icevis Schmidt, MS., show minor divergences from 

 the arrangements above, chiefly in the connections of the layers of 

 stellate spicules ; a similar divergence is shown by an unpublished 

 species of Tctliya from Mexico. 



Faringdon (Coral-Rag) Sponges.J — Mr. H. T. Carter draws 

 attention to some of the difiiculties which attend the view of Zittel 

 that these sponges are " Calcispongia^." He points out that their 

 resemblance to the genus Clathrina of Gray is only illusory, inasmuch 

 as what is hollow in the extant is solid in the fossil form. Tho 

 results of his earlier microscopic examinations were such as to lead 

 him to think that tri-radiatc spicules might have been present in the 

 coral-rag sponge, but further examination, while in some points 

 it appeared considc;rably to confuse matlcrs, seemed to provo that 

 there were no tri-radiatc siticulcs at all, and in a number of cases ho 

 found that the si)iculation of tho sjxmges was Lithistid, and his 

 anticipation that somo of the Faringdon forms were originally sili- 



♦ 'Zeitsch. wiss. Zool.,' xxxi. (1878) p. 68. 



t Ibid., xxviii.-xxxii. (various papers). 



X 'Ann. and Mag. Nat. Hist.,' iv. (1879) p. \M. 



