CHAP. VII.] THE DEEP-SEA FAUNA. 417 



As I have already said, the most remarkable new 

 forms are referable to the group which seems to be, 

 in a sense special to deep water, the Hexactinellidse. 

 I have already (p. 70) briefly described one of the 

 most abundant and singular forms belonging to this 

 order, Holtenia carpenteri ; and all the others, 

 though running through most remarkable variations 

 in form and general appearance, agree vrith Holtenia 

 in essential structure. In the Hexactinellidse all the 

 spicules, so far as we know, are formed on the hex- 

 radiate plan ; that is to say, there is a primary axis, 

 which may be long or short, and at one point four 

 secondary rays cross this central shaft at right angles. 

 Very often one-half of the central shaft is absent or 

 is represented by a slight rounded boss, and in that 

 case we have a spicule with a cross-shaped head, a 

 very favourite form in the manufacture, defence, and 

 ornament of the surface layer of these sponges ; and 

 often the secondary rays are undeveloped : but if 

 that be so, — as in the long fibres of the whisp 

 of Hyalonema, — in young spicules and in others 

 which are slightly abnormal, four little elevations 

 near the middle of the spicule, which contain four 

 secondary branches of the central canal, maintain the 

 permanence of the type. In many of the Hexac- 

 tinellidse the spicules are all distinct, and combined, 

 as in Holtenia, by a small quantity of nearly trans- 

 parent sarcode ; but in others, as in ' Venus's flower- 

 basket,' and the nearly equally beautiful genera 

 IpUteon, Aphrocallistes, and Farrea, the spicules 

 run together and make a continuous silicious net- 

 work. When this is the case the sponge may be 

 boiled in nitric acid, and all the organic matter and 



E E 



