riiijlogeny of the Ainpliidiscopliora. 4K;} 



(ii-sappear, and tlio icmaiiiini; axis would become vertically 

 orientated. 



Turning to the Ilcxasteroplioran tendency, we may suppose 

 that the body reticulum remained delicate and uniform and 

 did not assume a tendency to gather into strongly contractile 

 bands. Accordingly the slight centripetal |)re3sure of the 

 contractile strands would not have the etVect of bending back 

 the end rays^ which would grow outwards away from the 

 main rays, and, further, all six rays would be needed in a 

 uniform cubical — as distinct from laminate — reticulum. 



Although it would seem probable that a laminar condition 

 would arise out of a uniform cubical condition of dermal 

 reticulum, yet tiie Amphidiscophora, in several respects, would 

 appear to be more primitive than tiie Ilexasterophora. 



As Minchin * has pointed out, the collar-cell layer in the 

 former is nearer to the homocoelous state than in the latter. 

 Further, the skeleton is constructed on a simpler plan, i. e. of 

 a cubical framework formed from regular iiexactins, and, 

 lastly, the microliexactins remain in the condition of simple 

 hexactins. 



In the Hexasterophora, the hexasters have developed disks 

 of a third order at the ends of end rays, and there are no 

 true microliexactins, the so-called "derived hexactins" being 

 true hexasters, in which each main ray has one end ray. 



Leaving out of account for a moment the fact that the 

 aniphidisk is a diaster or didiscaster, the two great groups 

 might alternatively, but less aptly, have been named Micro- 

 hexactinophora and ilexasterophora : for in the first the 

 microscleres, apart from the amphidisks, are all hexactins, 

 and hexasters are never found ; in the second the microscleres 

 are all hexasters, and there appear to be no microliexactins. 



It might be supposed that pinular Hexasterophora (.-Lsfo- 

 neyna &c.) aie nearer to Amphidiscophora than are the noii- 

 pinular Ilexasterophora; even so, the pinular Hexasterophora 

 all have hexasters and are devoid of microliexactins. It is 

 conceivable that an Amphidiscophoran might lose its amphi- 

 disks ; its true character wouUl probably then be betrayed 

 by the presence of true microhexactins. 



In Amphidiscophora there would be, for obvious reasons, a 

 tendency to atrophy of the proximal ray of autodermal and 

 autogastral hexactins, thus giving rise to pinuli, for these rays 

 could not prevail against the concentric laminaj, but would be 

 suppressed or reduced to mere buttons, as we frequently see. 



Hexasterophora with pinuli {Asconema &c.) might be 



* E. A. Minchin, " A Speculation on the Phyloj^^eny of the Hexacti- 

 neliid Sponges," Zool. Anzeig. 1905, xxviii. pp. 439-448. 



