in POEIFERA 47 



oval ciliated blastula, with two cells at its posterior pole which are 

 interpreted as the mother cells of the archaeocytes or primitive ova. 

 From these cells numerous granular cells are budded off and fill the 

 interior of the vesicle, but other cells formed by the modification of 

 individual flagellated cells here and there, which lose their flagella, 

 also migrate inwards. These latter cells, at fixation, are stated to 

 burst forth and surround the ciliated cells. 



In Leucosolenia, on the other hand, the posterior part of the 

 interior of the blastula is filled with a mass of granular cells with 

 small nuclei, and in front of these is a tube of flattened pigmented 

 cells containing a lens-like body. This, according to Minchin, con- 

 stitutes, a rudimentary visual organ ; it disappears at fixation. The 

 cells of the posterior half of the blastula wall become granular in 

 situ during the free life of the larva, and so an arnphiblastula is 

 produced. (Fig. 25.) 



Further details of these interesting life-histories are urgently 

 called for. We wish to know what corresponds to the archaeocyte 

 in Grantia. If in this form the archaeocytes are only differentiated 

 from the dermal cells after fixation, this must surely be a more 

 primitive arrangement than what obtains in the Asconidae or in the 

 Tetractinellida, where these primitive ova are differentiated during the 

 segmentation of the egg. Minchin, indeed (1900), suggests that the 

 granular cells, which are invaginated whilst the embryo of this sponge 

 is in the tissues of the mother, are archaeocytes and are quite 

 distinct from the cells forming the one end of the arnphiblastula 

 which he regards as transformed flagellated cells ; but this view 

 is negatived by Dendy's researches, the results of which have been 

 described above. 



The development of the well-known freshwater sponge Spongilla, 

 which has been worked out in great detail by Evans (1899), presents 

 several features of great interest. This sponge belongs to the group 

 Demospongiae and forms a larva somewhat like that of Esperia, 

 but the outer flagellate layer extends all round. One end of the 

 larva is broader than the other and under this end is a cavity. The 

 rest of the interior is filled with yolk-bearing " archaeocytes," whilst 

 just under the skin is a layer of flattened cells with dense nuclei, 

 like those described in the interior of the larva of Grantia. It meta- 

 morphoses in much the same way as Esperia, i.e. it fixes itself by 

 the broad end ; but Evans maintains that some of the " flagellate 

 chambers " are formed at the expense of groups of archaeocytes and 

 do not owe their origin to the flagellated epithelium which invested 

 the surface of the larva. 



. Spongilla also reproduces itself by buds termed gemmules. 

 The development of these in the allied genus Ephydatia has been 

 worked out by Evans (1900). The gemmule first appears as a number 

 of wandering cells in the jelly of the sponge, which are distinguished 

 from their neighbours by possessing deposits of yolk in their cyto- 

 plasm. These cells gradually collect at a fixed point in the tissues, 



