Porifera. 



semi-viscous, semi-elastic Acarnus tortilis\ Microciona is similar but with smaller 

 elements. Sometimes in M. , more characteristically in Clathria coralloides and 

 Echinoclathria seriata, these cells are accompanied by long and very slender 

 elastic fibres , spiral when ruptured , stretched in every direction near the outer 

 and canal surfaces ; there are no terminal swellings as in the similar Hircinia 

 fibres (the nature of which T. considers not fully elucidated). Whether in Ectyo- 

 ninae they are formed by union of spherules, or are comparable to the sarcode- 

 strings of Halisarca and Bajalus, remains uncertain. 



Vosmaer & Pekelharing s chief part can be condensed in one line: - - Sollas's 

 membrane does not exist. Summary and criticism of the literature shows Delage 

 [cf. Bericht f. 1892 Porif. p 4] towards S.'s membrane an opponent in words, 

 believer in fact. Lendenfeld's opinions [cf. Bericht f. 1889 Porif. p 8] are in- 

 consistent with each other and with the facts admitted by all. S.'s statement that 

 the membrane is only found when flagella are not visible is xominous indeed . . . 

 the lines drawn by Dendy [cf. Bericht f. 1891 Porif. p 5] indicating S.'s mem- 

 brane are portions of flagella. A line could be found in such a position, thicker 

 than the collar outlines, careful focussing proved it a cylindrical body. Feeding 

 with carmine and milk did not support Bidder's hypothesis, the collars being 

 always separate , with long , distinct flagella. After one hour's feeding carmine 

 was found within the collars and in the lumen of the flagellated chambers as well 

 as in the collurn ; longer periods showed it first in the cell-bodies , then in the 

 pareuchyme. This strongly contradicts the theory of Bidder and Dendy that the 

 membrane acts as a filter. - - Flagella having been seen moving, portions of sponge 

 were placed one hour in osmic acid in the dark , then dialyzed into alcohol and 

 benzol ; in careless preparations the membrane was seen, never in well-preserved 

 tissue ; previous descriptions are probably based on ^portions of flagella and 

 collars, irregularly sticking to each other. Halichondria panicea, Spongilla 

 lacustris , and Sycon ciliatum were studied .- there was no membrane , nor direct 

 concrescence of the collars, nor any ground substance between them. 



Franze [vide supra Protozoa p 24] finds the collar in Choanoflagellata not 

 a closed cylinder or cone, but spathiform. 



2. Embryology. 



See also Maas i 1 ). 



Delage assures Maas [cf. Bericht f. 1892 Porif. p 5] that in Esperella, as in 

 the many other sponges observed, larvae leave by the osculurn : when issuing 

 through fractures they are often abortive. D. was right [cf. Bericht f. 1891 

 Porif. p 5] as to the cells which give rise to the adult ectoderm, there is only 

 possible doubt whether they are fundamentally distinct from cells of the central 

 mass. M's. intermediary cells are probably ciliate cells from which the flagellate 

 ends have been broken, his migrated cells at the anterior pole of the larva are prob- 

 ably drops of secretion. In M.'s own figures the true epidermic cells can be 

 recognized, their large nuclei are ascribed by him to the central mass. The sponge 

 does not as a whole execute amoeboid movements, M. has been deceived by the 

 protrusion and retraction of the marginal zone, the central part remains motion- 

 less. D. complains that M. has not acknowledged though he has confirmed D.'s 

 discoveries [v. Bericht f. 1891 Porif. p 5]. Ensomme, deux faits nouveaux 

 sont venus modifier du tout au tout la conception ancienne des Eponges siliceuses 

 et de leur developpement. Ce sont la formation de 1'^piderme aux d^pens de cel- 

 lules primitivernent interieures, et celle des corbeilles aux de'pens des cellules cilie'es 

 rentrees a I'interieur. Pour Tun comme pour 1'autre, je reclame la priorite". 



