43 



It should be added that Mr Rousselet prefers formalin to glycerine as a medium for 

 mounting; but I have myself used glycerine on the ground that there is less risk of drying up 

 if the cementing process has not been completely successful. 



The arrangement of the genera of non-encrusting Ctenostomata is admittedly a matter 

 of great difficulty; and they have been groupecl in a considerable variety of ways by different 

 authors. Waters ] ) has pointed out that the Division Stolonifera, as used by rnany authorities, 

 includes more than one type of structure, and that its type-genus, as originally introduced by 

 Ehlers, is Hypophorclla. He accordingly suggests the following classification : — 



Vesicularina. "Usually 8 — 10 tentacles ; a gizzard; the zooecia grow direct from the stem and 

 "have no independent movement". Genera: — Zooóotryou, Bowcrbankia, Amathia, 

 Vesicttlaria, ? Avenella, Buskia (B. soda/is), Cryplopolyzoon. 



Stolonifera. "There is a delicate creeping rhizome, which at intervals expands, and from 

 "these expansions the zooecia arise usually in pairs; no gizzard". Genera: — 

 Hypophorella, Farrclla, Valheria, Mimosella, Triticella, Cylindroecium, Buskia 

 (B. setigerd). 



Waters is uncertain where to place Arachnidium (which may be allied to Cyliiidroeciitiu), 

 Victorella and Arachnoidea. 



Rousselet 3 ) has proposed the association of the fresh-water genera Paludicella, Victo- 

 rella, Pottsiclla and Arachnoidea, with the marine genera Arachnidium and Cylindroecium 3 ), 

 under the name of "Cruciform Stolonifera"; for which the name Paliidicellea, already in use 

 with a more restricted meaning, might be employed. The study of the 'Siboga' material seems 

 to me to lend much support to this grouping, which is based on the fact that each zooecium 

 typically gives off a terminal daughter-zooecium, and another on each side, more or less at 

 right angles to the parent; thus producing a cruciform arrangement of uniserial zooecia. The 

 term "stolon", which is sometimes applied to the narrow tubes which connect the expanded 

 parts of the zooecia, is not very suitable in this connexion, since the tubes in question are 

 merely the proximal parts of the zooecia themselves. 



Group B. Paludicellea. 



Paliidicellea Allman, 1856, «Mon. F.-W. Pol." Ray Soc, pp. 10, 76. 

 Paludicellina Jullien, 1S88, "Miss. Sci. Cap Horn", VI, «Bry."', p. 7. 

 Paludicellina Annandale, 191 1, "Faun. Brit. IncL", "Freshwater Sponges . . . and Pol.", p. 190. 



In this division may provisionally be associated the Families Paludicellidae, Victorellidae, 

 Arachnidiidae and Xolellidae (= Cylindroeciidae, auctt.), the first of which does not come within 

 the province of this Report. 



1) Waters, A. W., 1910, pp. 23S, et seq. 



2) Rousselet, C. F., 1907, Proc. Zool. Soc, p. 252. 



3) Described as KoUlla in the present Report. 



43 



