DENDY— HOMOSCLEROPHORA AND ASTROTETRAXONIDA 



250 



cladi occasionally bifurcate at the apex. In a typical example the shaft measured about 

 0'2 by 0"03 mm., cladi about the same. The forms with slender rays ai-e perhaps young. 



(2) Oxea (fig. 5 b) ; fairly stout, fusiform, slightly curved and gradually and 

 sharply pointed ; varying occasionally to stylote (fig. 5 h') or even strongylote (fig. 5 h") 

 forms; often slightly centrotylote ; typical size about 07 by 0'025 mm. 



(3) Aspidasters (fig. 5 c — 5 c"") ; thin, plate-like, oval ; usually with fairly even 

 outline ; surface beset with numerous, scattered, slightly stellate, short spines or tubercles ; 

 size about 0"15 by 0'07 mm. The young forms are very thin, smooth, oval plates, with a 

 marginal fringe of slender spines ; radially striate when very young. 



(4) Chiasters (fig. 5 d) ; polyactinal strongylasters and tylotasters ; about 0"012 mm. 

 in total diameter ; sometimes with a distinct centrum. 



(5) Microrhabds (fig. 5 e) ; smooth, fusiform, usually slightly curved, usually sharply 

 pointed, often centrotylote ; size about 0"053 by 0"004 mm. Abundantly scattered in the 

 choanosome and in the pore-sphincters. 



With regard to the microscopical anatomy of the soft parts in this form I have only 

 to note two points. The inhalant pores are situated each in a distinct gap in the dermal 

 armour of aspidasters, and each is provided with a thin sphincter membrane strengthened 

 by numerous microrhabds. The interior of the sponge is charged with numerous small, 

 brown, granular pigment-cells, especially abundant near the surface. 



This species comes very near to Erylus carteri SoUas [1888], originally described by 

 Carter from the Gulf of Manaar under the name Stelletta euastrum [1880 B, p. 135], which 

 was also associated with Siliquaria. The most important difference lies in the fact that 

 the microrhabds (microxea) are smooth, as usual in the genus, instead of minutely spined 

 as described and figured by Carter for the type of the species. 



To judge from the emended description given by Sollas [1888], Kidley's [1884 c] 

 Erylus cylindrigerus from the Mascarene Islands is perhaps even more closely related to 

 the present species, differing, however, in the lozenge-shaped form of the aspidaster, the 

 form and size of the small asters, and the (constant ?) strongylote form of the main skeleton 

 spicule. It appears to me highly probable that all three species will have to be united in 

 the futui-e. 



Register No., Locality, (&c. LVii. 7, Cargados, 30 fathoms. 



Family Donatiidse 

 = Tethyadm or TetJiyidce auctorum. 



Astrotetraxonida without tetractinellid megascleres. With a strongly developed 

 fibrous cortex. Main skeleton composed of radially arranged styli (or oxea ?). Microscleres 

 euasters of various forms, including large spherasters, to which microrhabds may be added. 



This family appears to have originated from some corticate stellettid ancestor of 

 the Aurora type. Indeed, if we admit such " epipolasid " forms as Aurora distinctus, 

 A. cribriporosa and A. sterrastrcea amongst the Stellettidse, it is not easy to frame a 

 diagnosis by which the Donatiidse can be logically excluded from the same family. 



33—2 



