EUPLECTELLA. 45 



While in certaiu species the sj^icules of the skeletal frame- 

 work remain perfectly separate throughout life (f. i., in E. marshaUi, 

 oweni, curvistellata), in others they begin at a certain age to 

 undergo fusion by means of syuapticulse. The soldering together 

 is nearly exclusively confined to the beams above described and 

 to their direct continuations into the sieve-plate. The process 

 seems always to commence at the lowest base of the sponge-wall 

 and to proceed gradually upwards, either to stop at some point 

 on the sides (f. i., E. imperialis) or to extend up into the sieve- 

 plate beams {E. aspcvgUIum). In view of the fact that in so 

 many sessile Lyssacina exactly the same fusion of spicules first 

 takes place where the surface is in contact with the substratum, 

 I am led to the assumption that the latter exercises a certain 

 influence in inducing the soldering process in question. The cases 

 of Euphclella species, in which the soldering never takes place, 

 may be explained by the fact that in them the lowest end of the 

 body-wall is not in direct contact with, but stands above, the 

 bottom-surface. That the basal tuft never becomes ankylosed, 

 though penetrating quite into the substratum, is evidently due 

 to its being devoid of living soft parts. 



Apart of the above framework, the main mass of the 

 choanosome is supported by numerous other parenchymalia of 

 variable size and strength (principalia and accessoria), arranged 

 either loosely in indefinite order or grouped into strands run- 

 ning in various directions. The forms of individual spicules are 

 here again predominantly thetactins ; less frequently hexactins, 

 paratetractins and diactins, and rarely pentactins or stauractins. 

 These constitute the principal part of the so-called flake-tissue or 

 ' Flockengewebe ' investing the skeletal latticework of the lateral 

 wall on the external side and forming the main mass of the ledges. 



