44 M. E. MetschnikofF on the 



From the preceding it follows that the two principal layers 

 of the sponge-body are founded already in the body of the 

 larva, and, further, that the upper ciliated half of the body is 

 converted into the entoderm, and the lower unciliated half into 

 the layer surrounding the entoderm. 



Having now communicated the facts of my investigations, 

 the question may be put, How far can the results obtained 

 by me be brought into accordance with Hackel's statements ? 

 At pp. 34 and 216 of his work this naturalist gives the fol- 

 lowing short summary of the developmental phenomena in 

 the Calcispongise: — "From the egg is produced, in consequence 

 of total regular segmentation, a simple spherical or elongated 

 round body, which is at first composed of homogeneous spherical 

 cells. Then there is produced in the interior of the cell- aggre- 

 gate a small central cavity (the stomach), which, breaking- 

 through outwards, forms an orifice (the osmium or buccal 

 orifice). The surface becomes covered with cilia ; and then 

 the embryo swims about as a free larva [planida) for a long 



time The body-wall (of the larva;) consists of two 



layers of cells, entoderm and exoderm. The inner layer, or 

 entoderm, consists of a layer of unciliated cells ; the outer 

 layer or exoderm consists of a layer of ciliated cells (flagellate 

 cells)." Then "the larva falls to the bottom, and attaches 

 itself. The attachment takes place at the pole of the longi- 

 tudinal axis, opposite to the buccal orifice (aboral pole), by a 

 flat or peduncular surface of adhesion, which from this time 

 forth forms the base of the sponge-body. The flagellate cells 

 of the exoderm now retract their flagellar filaments, coalesce 

 to form the syncytium, and begin to secrete their interior pro- 

 toplasmic products, the calcareous spicules. The cells of the 

 entoderm, on the contrary, which were previously not ciliated, 

 stretch forth each a long vibratile process, and thenceforward 

 line the surface of the stomach as a flagellate epithelium." 



This description therefore runs quite differently from that 

 given by me above ; for according to Hiickel the skeleton- 

 forming layer (exoderm, Hack.) originates not from the un- 

 ciliated cells, but from so-called flagellate cells provided with 

 long cilia ; and the converse is the case with the entoderm, 

 which Htickel derives from spherical cells, whilst, according 

 to my observations, it takes its origin from the ciliated (or 

 flagellate) cells. This is the chief difference in our state- 

 ments, the elucidation of which will here be attempted. 

 Hackel describes the larva? of four species of Calcispongiae, 

 of which those of Sycyssa Huxleyi exhibit the greatest analogy 

 with the larva 1 of Sycon, inasmuch as they appear to be com- 



