182 



tes that their function is excretory and that they are provided with a 

 fine duct opening on the surface. I have never seen any structure 

 in them resembling a duct. Von Lendenfeld (3) describes the 

 nucleus of the ectoderm cells (p. 398) as »brotlaibformig« and in the 

 contracted ectoderm-cells of Sycandra raphanus figures two nuclei 

 (Taf. XIII Fig. 102), a spherical one at the base and another »brotlaib- 

 förmig« in the upper expanded portion; and in this manner he makes 

 two cells, a supposed mesodermal one with the spherical nucleus and 

 a flattened ectoderm cell with the other nucleus. As a matter of fact 

 his descriptions and figures of the ectoderm cells are throughout erro- 

 neous and the contracted ectoderm-cells have not two nuclei but one 

 only the spherical one at the base, which is the true ectoderm nucleus. 

 Where the sponge is in contact with the foreign body, whatever it may 

 be, on which it is growing, the ectoderm cells are very columnar, gra- 

 nular, and of irregular outline, and are no doubt glandular, forming 

 some secretion by which the sponge adheres to the base on which it 

 grows. 



2) The Pores. At the ends of growing branches every stage in 

 the formation of pores can readily be found , though they also occur 

 almost everywhere in the sponge. 



The first stage is an ectoderm cell which is somewhat more gra- 

 nular than the ordinary cells , with protoplasm of a characteristic 

 yellowish brown tinge, and having distinct cell limits. Such a cell 

 then grows inwards towards the endoderm, and reaching it, pushes its 

 Avay between the collar cells, while still retaining a connection with 

 the remaining ectoderm. The nucleus , which does not alter in any 

 way , may remain at the level of the ectoderm , or may grow in with 

 the cell and be found at any level. Pores in this stage are very com- 

 mon in sections, and have led to much error, since the cell is very large 

 and often grows inwards in a slanting direction, and its connection 

 with the outer surface is always very delicate. Hence the cell may 

 appear in sections as if lying entirely between the endoderm cells, or 

 entirely in the mesoderm , or partly in one , partly in the other posi- 

 tion. It is these cells which I (4) formerly wringly described as amoe- 

 boid mesoderm cells (p. 265, PI. XI, Fig. 22) while von Lenden- 

 feld (3) , finding them in the endoderm, has without further investi- 

 gation described them as «Kragenmutterzellen«, and a similar position 

 has apparently caused Bidder (2) to attribute an endodermal origin to 

 the pores. After the cell has reached this stage, it spreads out and 

 becomes perforated. The fully formed pore is a single cell with a nu- 

 cleus exactly similar to the remaining ectoderm nuclei and an intra- 

 cellular duct, which has a wide inner opening and a very delicate 



