EUPLECTELLA. 39 



marshalli (PI. IV, fig. 0). Moreover, it is persistent in the genus 

 Holascus. 



That ground-form, in order to differentiate into Eupledella, 

 essentiall}'- needs but to open on the lateral wall a number of 

 additional oscula, which might conveniently be called the 'parietal 

 oscula in distinction from the more primary oscules (i.e., the meshes 

 of the sieve-plate). The parietal oscula (Derraalostien Marshall, 

 parietal gaps or Wandlücken F. E. Schulze) are round, because 

 isolated, unlike those of the sieve-plate, which, being crowded 

 together, are more or less angular. It is interesting to note in 

 this connection that in a certain Eiiplectellid {Walteria flemmingi) 

 the entire lateral wall presents an appearance quite like that of 

 the Euplectella sieve-plate. The parietal oscula are surrounded 

 by a narrow, iris-like, circular membrane, — the oscular membrane, 

 — formed by the confluence of cobweb-like trabeculae at the 

 edge where the external and internal surfaces join. The same 

 membranous edge may sometimes be observed, though much less 

 conspicuously, at certain parts of, if not all around, the sieve- 

 plate meshes. The assumption that the parietal oscula can be 

 closed by the activity of the tissues during life (Marshall '75, 

 p. 197) is, in my opinion, without foundation and highly im- 

 probable. They are certainly not provided with a tissue which 

 is any more contractile than the trabecula3 ; let alone then a 

 definitely developed sphincter muscle. 



In my estimation, the parietal oscula are to be collated with 

 those secondarily formed oscules, which are so commonly met 

 with, to the utter confusion of the question of individuality, in 

 all groups of the Spongida. Eufpleciella is then rather polyzoic 

 than monozoic, if it be necessary to use such qualifying terms 

 at all. In this genus, as also in Regadrelh, Tcegeria, &q., the 



