Hexactine Spicule of ILxactinelliJa. 507 



Tlie cliaracferistic feature of the Ilcxactinellid sponge 

 wliirh must liavo existed bot'ore stauractine or hcxactine 

 spicules arose, ami which pmbahly conditioiio 1 the shape of 

 those, spicules, is the dcrtnal* .syncytial network. 



if purely sfanractiiie spon;;es existed, it was because this 

 network woidd be extremely thin at first, and Nature would 

 liave the problem of forminf^ a practically — not, of coun=e, 

 mathematic;illy — two-dimensional sca(loI(iin<x (/. e. staur- 

 actine). When in course of time the network grew tliioker, 

 a three-dimensional scaflohiing would become necessary. 



It has been supposed that the biological coiulitions, which 

 would account for the existence of the regular hexaetitie form, 

 do not occur, because the meshes of the trabecular network 

 are of all shapes. It seems to mc that these corxlitions do 

 actually exist. Nature has a very elusive material to deal 

 with in the ense of the fluent and contractile sym^ytium of 

 the Hexaclinellid sponge, and it would be impracticable to 

 construct a scaffolding that would exactly folio rv the protean 

 form of a syncytial network. 



In the Hexaclinellid sponge, it is not necessary to look 

 for a perfectly regular network of protoplasmic strands to 

 account tor a similarly regular network of supporting scat- 

 folding ; the protoj)lasmic strands of syncytium need only 

 loosely drape the supporting rods ; they may fill in the sharp 

 corners, form subsidiary meshes in the cubical compartments, 

 anil lastly prod out the walls of the cubicle with spanrlrils or 

 end rays. For the su|)port of three-dimensional spaces of a 

 network, the most economical and etfii:ient scaffolding is the 

 cubical one. The selected sclerite has bi;en one with si.x ravs 

 at right angles, giving support in the direction of length, 

 breadth, and depth. 



Nature, having selected the hexactin, has certainly stereo- 

 typed her pattern in a wondertul manner. It is always a 

 surprise to see the axis-cross persisting in long slender 

 diactine and even in monactine spicules, such as the distil 

 pronged knob of root-tuft spicules of Hyalonema. No 

 wonder we are led to account for such phenomena by calling 

 in tlie mysterious moltcular forces of crystallisation in place 

 of, or in aid of organic forces acting en vvisse. 



We find, however, that Nature readily adapts her methods to 

 the changing requirements. As I have already endeavoured to 

 provet, when it becomes a question of supporting concentric 

 lamii:a3 of reticulum, the two tangential axes essential in a 



* Dermal, as contrasted with gastral, and inchiding dermato3ouie and 

 jjastrosome. 



t Ann. Mnjr- N. H. 1909, (viii.) vol. iv. p. 179. 



