B. Porifera incalcaria. 2. Cornacuspongiae. 7 



flattened ; the flagellated cells lose (retract) tlicir flagellum, become irregularly 

 shaped, and less regularly placed: les unes s'enfoncent, les autres restent a leur 

 niveau et toutes s'ecartent quelque peu de leur voisines. Meantime the epidermic 

 cells come to the periphery. There they grow broader and their limits stick to- 

 gether (se soudent). The flagellated cells then disperse more and more through 

 the whole sponge-mass and begin to be captured by the amoeboid cells [cf . Bericht 

 f. 1891 Porifera p 5]. The sponge now is totally covered by the epidermis, while 

 the inner mass consists of amoeboid cells with the captured flagellated cells, and 

 the intermediate cells. This cell-mass leaves small lacunae between, thus forming 

 a network with meshes. Towards the beginning of the second day the amoeboid 

 cells (now called groupes polynuclees) begin to swell , because of the extension 

 of the devoured cells. The meshes of the network sometimes fuse, forming larger 

 lacunae: the beginning of the exhalant system. At the same time some of the 

 groupes polynuclees come nearer to each other; their flagellated cells are arranged 

 around the meshes , forming a lining of a cavity : the beginning of a flagellated 

 chamber, which soon enters in communication with an exhalant cavity. The latter 

 is afterwards lined by flattened intermediate cells. At about the same time that the 

 flagellated chambers are formed, appear pores and an osculnm. 



Zykoff ( 2 ) states some points in the development of the gemmulae ofEphydatia 

 fluviatilis. In the ordinary amoeboid cells refringent corpuscles (glanzende Korn- 

 chen) appear. These cells, then called Trophophoren, begin an einander zu 

 gleiten, unite together and form small spheres around which other cells of the 

 parenchyme lay themselves in several layers. The arnphidiscs do not develop 

 in the peripheral cells , but externally from this layer. Those keulenforrnige 

 peripheral cells excrete a chitinous sheet. 



WiGPZejski mentions the occurrence ofCarierius stepanovii Petr and Heteromeyenia 

 repens Potts in Galicia. Probably Ephydatia bohemica Petr is a variety of C. ste- 

 panovii. There is no sharp distinction between freshwater sponges the gemmulae 

 of which possess amphidiscs, and those with Belegnadeln<r. 



Zykoff ( 3 ) studied the development of Ephydatia millleri Lbk. from gemmulae 

 which he had kept dry for two years. A fortnight after seeding them, they began 

 to develop. The young sponge consists of a mass of amoeboid cells covered with 

 epithelial cells. Two days later spicules appear ; internally some lacunae, lined 

 with epithelium, are present and soon afterwards an osculum is visible. Flagella- 

 ted chambers appear afterwards; first as solid cell-masses, later as hollow sacs. 



Hornell relates a case of commensal ism between Microciona plumosa Bwk. 

 and Leucodore coeca Oerst. 



B. Ceratina. 



See, supra p 4, Topsent( 2 ). 



Lendenfeldp) divided [cf. Bericht f. 1889 Porifera p 7] the siliceous sponges 

 into two subclasses, viz. Triaxonia andTetraxonia. The former group has two or- 

 dines, viz. Hexaceratina and Hexactinellida. The author now describes the 4 species 

 of sponges belonging to the Hexaceratina which occur in the Adriatic. Darwinella 

 aurea Fr. Mull, is more fully described than in the Monograph of the Horny Spon- 

 ges.* Externally the sponge resembles much Aptysilla sulfurea F. E.S., especially 

 as regards the distribution of the incurrent apertures, the conuli, the oscula, etc. 

 The canal system differs likewise little from that of A. s. ; even those wide canals 

 between the groups of flagellated chambers, which the author considers something 

 very remarkable in A., seem to occur sometimes in D. [The author does not 

 point out this great resemblance.] As to the skeleton, in both species we find a 



