SPONGES. 225 



specific identity. In the Ceylon specimens, however, there appears to he (usually af 

 any rate) no sand in the skeleton fibres. Keller makes his species the type not 

 only of a new genus, but even of a new family. He considers it to be closely related 

 to Aplysillu, but, unfortunately, his material did not enable him to investigate the 

 form and arrangement of the flagellate chambers, or he would probably have seen 

 that the affinity was rather with Aplysina, as the firm, almost leathery texture of 

 the living sponge and its stony hardness when dry might alone have indicated. I 

 follow Mr. Carter in retaining the Ceylon species, at any rate (with which Keller 

 does not appear to have been acquainted), in the latter genus, from which it differs in 

 no important respect. In any case the name Psammaplysilla appears to have been 

 very unfortunately chosen, as the sandy character of the fibre is hardly of generic 

 value, and the relationship with AphjsiUa is not nearly so close as that with Aplysina, 

 a very distinct genus. 



R.N. 95 (Stat. IV., off Karkopani, 6-9 fithoms, Gulf of Maiiaar). 



Aplysina herdmani, n. sp. Plate XVI., fig. 4. 



The single specimen consists of a rather thin, irregular, flattened crust, from which 

 short, slender, cylindrical, digitiform processes rise vertically upwards at wide and 

 irregular intervals. The ends of these processes are truncated, and each has a single 

 small vent in the middle.* The surface, both of the basal crust and of the dicitiform 

 processes, is glabrous, hut beset with numerous minute conuli, from the apices of 

 which the ends of the primary fibres sometimes project. The colour, in spirit, is dull 

 purple throughout ; the texture compact and rather fleshy, but compressible and 

 resilient. The maximum diameter of the basal crust is about 55 millims. and its 

 thickness about 4 millims. The digitiform processes are about 11 millims. high by 

 only 2 millims. or 3 millims. in diameter. 



The skeleton, in the basal crust, consists of a reticulation of thin-walled, pithed 

 fibres of a rather dark brown colour, amongst which distinct primary fibres, running 

 vertically into the surface conuli, are clearly differentiated (Plate XVI., fig. 4). These 

 primary fibres are about 0"08 millim. in diameter, and the very thick, granular 

 " pith " contains abundant broken sponge spicules as foreign inclusions. The 

 primary fibres sometimes branch, and they are connected together by a network of 

 secondaries which vary greatly in diameter, being sometimes as stout as the 

 primaries and sometimes very slender, only about 0'016 millim. in diameter. The 

 secondary fibres are free from foreign matter and the stouter ones commonly run 

 across between the primaries and thus form rectangular meshes, but the meshes are 

 usually irregularly polygonal and very variable in diameter. 



In the digitiform processes the main fibres containing broken spicules run longi- 



* One of the processes forks into two close to its extremity, and each of the very short brunches thus 

 produced bears a small vent on its truncated end. 



2 G 



