Sponge-fauna of Norway. 163 



sence of the latter, so chavacteristic tliroiigliout the Des- 

 macidina, wonkl of itself have afforded us a hint as to the 

 alliances of Tetilla ; but, taken in conjunction with the evi- 

 dence furnished by the ends of the spicules, it gives us a very 

 strong case indeed. The swollen terminations of the spicules 

 of ^. sclumns suggest the possibility of a polyphyletic origin 

 for the TetractincUida. The trichite sheaves of llhaphido' 

 theca deserve a word of mention ; for though they are found in 

 several grou]3s of sponges, yet they are most commonly present 

 in Desmacidina and Stelletta] and a genetic connexion be- 

 tween these two groups being probable on other grounds, we 

 may regard the sheaves as derived from a common ancestor, 

 and thus gain some hoipc of tracing out their origin in the 

 other group of sponges in which they occur. 



If the Tetilla embryo, taken in conjunction with Bhapltido- 

 iheca^ furnishes evidence of a passage from the Monaxinellida 

 to the Tetractinellida, so no less does it show by the late 

 development of its rind a passage from the non-corticate to the 

 corticate sponges, and proves, what we should have ])redicted 

 on a 'priori grounds, that the latter were preceded by the 

 former. Schmidt's group of Corticata (Rindeuschwiimme), 

 though heterogeneous as at iirst constituted, appears to me a 

 good one if restricted to Tetractincilid sponges possessing 

 rinds ; and I consequently adopt it so amended in the table 

 which follows below. This appears to be a good place for 

 offering a few observations on the classification of the Tetrac- 

 tinellida. In my paper on the Trimmingham Ihnts I pro- 

 posed to divide the Tetractinellida into those with loose or 

 separate spicules (Choristidai) and those with spicules mutu- 

 ally interlocked to form a network (Lithistidas). Zittel, 

 placing greater stress on the branching character of the latter 

 spicules, had previously declared for a wider separation between 

 the Choristidai (Tetractinellida) and Lithistids than this would 

 imply; but Oscar Schmidt (rightly, as it seems to me) declares 

 this four-rayed character of the spicules to be fundamental, 

 although he maintains the ordinal distinction of Tetractinellida 

 (my (Jhoristidse) and the Lithistidai. The presence in the 

 Lithistidte of trifid forks and anchors precisely similar to those 

 of the Choristida3 is for me a strong point in favour of their 

 common derivation ; and the primary difference lies in the 

 different character of the chief four-rayed elements. In the 

 Ghoristidai the chief tetractincilid spicules are confined to 

 the surface (the tetractincilid character is only skin deep) ; and 

 they are clearly differentiated into shaft and rays, which liave 

 a very definite direction with regard to the surface of the 

 sponge, the shaft being usually radial and the rays developed 



